GgEEc Ieik
V gesourxe vny Energy Effixient
cetwork for Ieik
FednanVa M. H. Ramae
Clade Hall
Unihedeify af CambdiVge
A fheeie egbmiffeV fad fhe Vegdee af
DocPor of PDilosopDy

Dexlvrvtion
This dissertation is the result of my own work and includes nothing which is the
outcome of work done in collaboration except where specifically indicated in the text.
This dissertation does not exceed the regulation length of 60 000 words, including
tables and footnotes.

To Susana and Joa˜o Tiago, the loves of my life.

Vxknowleygements
First and foremost, I would like to acknowledge my supervisors, Professor
Jon Crowcroft and Professor Ian White. Their support has been invaluable.
Without their advice and superior supervision I would have accomplished
little, if anything. I have especially appreciated their encouragement to
pursue a wide range of research topics and the total freedom conceded to
develop my ideas. I am therefore truly thankful to them.
I gratefully acknowledge the financial support of the Foundation for Science
and Technology, from Portugal, without whose financial support it would
not have been possible for me to undertake the research leading to this
dissertation.
Considerable thanks is also due to some people I had the privilege to work
with very closely: Richard Gibbens, Pablo Rodriguez, Fei Song, Alessio
Giorgetti, Xavier Amatriain. Thank you for all the thought-provoking sug-
gestions and ideas.
The work described here could not have proceeded satisfactorily without the
excellent support I received from the Computer Laboratory, the Department
of Engineering, and the University of Cambridge as a whole. In particu-
lar, I have to emphasise the importance of the discussions, the sharing of
ideas and the feedback I received from the Networks and Operating Sys-
tems group (NetOS). Thanks to Eiko Yoneki, Steven Hand, Andrew Moore,
Philip Watts, Derek Murray, Chris Smowton, Tim Griffin, Cecilia Mascolo,
Vaughan Wittorff, Stephen Kell, David Miller, Nishanth Sastry, and to all
my other colleagues from NetOS I do not have the space to mention here.
I would also like to thank all my other colleagues and office mates from
the Computer Lab, with whom I spent a great time, playing table foot-
ball, drinking lots of beer and happily procrastinating: Amit, Fei, Periklis,
Narseo, Bharath, Kiran, Tassos.
Besides my friends from the lab, I made other very good friends for life
in Cambridge. They made the time spent in this beautiful city even more
worthwhile. Thank you to Christoph, Jin, Dominik, Susanne, Maurizio,
Sabine, Mathias, Mo, Maria, Pavlos. Thank you for all those great discus-
sions and those great outings to the pub. Cheers, mates!
I have to acknowledge also all the people from my college, Clare Hall. They
made me feel at home. The time I spent as a member of the Graduate
Student Body was particularly rewarding and fun. Thanks also to my
housemates. I miss those amazing dinners trying awkward food (for my
particular taste of course!) and delightful wine from all over the world.
Thanks to Cambridge, a truly inspiring and vibrant city. Thank you to The
King Street Run, to The Eagle, to The Castle, to All Bar One, and to the
dozens of pubs in Cambridge, London, and all around the UK where I have
drunken pints and pints of beer of countless styles and a myriad of flavours.
As part of the PhD I was fortunate to spend several months as a research
intern in Telefonica Research Barcelona. Thank you to Barcelona, one of
the most pleasant and diverse cities in the whole world, and thank you to all
the people that made my life and my work there so fulfilling. In particular,
thanks to Pablo, Xavier, Nico, Alberto, Georgos, Staluta, and Xiao. Also
a very warm thank you to my good friend Andreas, who made our stay in
Barcelona sincerely unforgettable.
Although living in Cambridge and in Barcelona was great, it was very hard
being away for so long from all my friends in Portugal. I will not mention
any names, as unwittingly forgetting a single one would be inexcusable.
But I thank you all for your friendship and care. Albeit far, I felt you were
always close.
As hard as it was to be away from my friends, it was even harder to be
away from my loving family. Thanks to my Mum, Dad, and Grandparents,
and also to the rest of the family for their endless support. Thanks also to
Dalinha, Quim, Rute, and Ricardo for accepting me in their family and for
treating me so amazingly well.
Finally, I would like to thank the most important people in my life. Joa˜o
Tiago, you are only months old and you’ve already changed my life so com-
pletely! Thank you so much for being the nicest and most beautiful baby
in the universe. Thank you in particular for your smiles - they make my
day, every day. Very special love and a huge thank you goes to my wife,
Susana, the person I definitely owe the most. Susana, you have endured
much undeserved neglect to allow me to concentrate on my work. But I
received from you unconditional support at all times. You gave me peace
and tranquility. Inspiration. Friendship and love. You were always, always,
always here for me. No words could ever express my gratitude. Thank you
for being who you are, and for loving me.
dwrigvyo.

GREEN IPTV: A Resource and Energy Efficient Network for IPTV
Fzrnvnyo bC kC gvmo“
Vwstrvxt
The distribution of television is currently dominated by three technologies:
over-the-air broadcast, cable, and satellite. The advent of IP networks and
the increased availability of broadband access created a new vehicle for the
distribution of TV services. The distribution of digital TV services over IP
networks, or IPTV, offers carriers flexibility and added value in the form
of additional services. It causes therefore no surprise the rapid roll-out of
IPTV services by operators worldwide in the past few years.
IPTV distribution imposes stringent requirements on both performance and
reliability. It is therefore challenging for an IPTV operator to guarantee
the quality of experience expected by its users, and doing so in an efficient
manner. In this dissertation I investigate some of the challenges faced by
IPTV distribution network operators, and I propose novel techniques to
address these challenges.
First, I address one of the major concerns of IPTV network deployment:
channel change delay. This is the latency experienced by users when switch-
ing between TV channels. Synchronisation and buffering of video streams
can cause channel change delays of several seconds. I perform an empiri-
cal analysis of a particular solution to the channel change delay problem,
namely, predictive pre-joining of TV channels. In this scheme each Set Top
Box simultaneously joins additional multicast groups (TV channels) along
with the one requested by the user. If the user switches to any of these
channels next, switching latency is virtually eliminated, and user experi-
ence is improved. The results show that it is possible to eliminate zapping
delay for a significant percentage of channel switching requests with little
impact in access network bandwidth cost.
Second, I propose a technique to increase the resource and energy efficiency
of IPTV networks. This technique is based on a simple paradigm: avoiding
waste. To reduce the inefficiencies of current static multicast distribution
schemes, I propose a semi-dynamic scheme where only a selection of TV
multicast groups is distributed in the network, instead of all. I perform an
empirical evaluation of this method and conclude that its use results in sig-
nificant bandwidth reductions without compromising service performance. I
also demonstrate that these reductions may translate into significant energy
savings in the future.
Third, to increase energy efficiency further I propose a novel energy and
resource friendly protocol for core optical IPTV networks. The idea is
for popular IPTV traffic to optically bypass the network nodes, avoiding
electronic processing. I evaluate this proposal empirically and conclude
that the introduction of optical switching techniques results in a significant
increase in the energy efficiency of IPTV networks.
All the schemes I present in this dissertation are evaluated by means of
trace-driven analyses using a dataset from an operational IPTV service
provider. Such thorough and realistic evaluation enables the assessment
of the proposed techniques with an increased level of confidence, and is
therefore a strength of this dissertation.
aist of puwlixvtions
In the course of my studies I have published the papers and technical reports presented
below. Some papers discuss topics covered in this thesis, while others describe distinct
research threads. The paper “Channel Smurfing: Minimising Channel Switching Delay
in IPTV Distribution Networks” has been given a best paper award.
[2012] H. Kim, J. Crowcroft, and FC bC kC gvmos. Efficient channel selection
using hierarchical clustering. In lolbob, San Francisco, CA, June 2012.
[2011] FC bC kC gvmos, J. Crowcroft, R. J. Gibbens, P. Rodriguez, and I. H.
White. Reducing channel change delay in IPTV by predictive pre-Joining of TV Chan-
nels. hignvl eroxz““ingO Imvgz Xommunixvtion, 26(7):400412, 2011.
[2010] FC bC kC gvmos, R. J. Gibbens, F. Song, P. Rodriguez, J. Crowcroft, and I.
H. White. Reducing energy consumption in IPTV networks by selective pre-joining of
channels. In hIGXdbb fiork“hop on grzzn nztfiorking, New Delhi, India, Aug. 2010.
[2010] FC bC kC gvmos, J. Crowcroft, R. J. Gibbens, P. Rodriguez, and I. H.
White. Channel smurfing: Minimising channel switching delay in IPTV distribution
networks. In IXbE, Singapore, July 2010.
[2010] F. Song, H. Zhang, S. Zhang, FC bC kC gvmos, and J. Crowcroft. Relative
delay estimator for SCTP-based Concurrent Multipath Transfer. In GadWEXdb,
Miami, FL, Dec. 2010.
[2009] FC bC kC gvmos, F. Song, P. Rodriguez, R. Gibbens, J. Crowcroft, and
I. H. White. Constructing an IPTV workload model. In hIGXdbb po“tzr “z““ion,
Barcelona, Spain, Aug. 2009.
[2009] FC bC kC gvmos, A. Giorgetti, F. Cugini, P. Castoldi, J. Crowcroft, and
I. H. White. Power excursion aware routing in GMPLS-based WSONs. In dFX, San
Diego, CA, Mar. 2009.
[2009] F. Song, H. Zhangy, S. Zhangy, FC bC kC gvmos, and J. Crowcroft. Relative
delay estimator for multipath transport. In XocEmi htuyznt lork“hop, Rome, Italy,
xii
Dec. 2009.
[2009] F. Song, H. Zhang, S. Zhang, FC gvmos, and J. Crowcroft. An estimator of
forward and backward delay for multipath transport. Xomputzr avworvtory izxhnixvl
gzport jXVbBXaBigBLIL, March 2009.
[2008] FC gvmos. Design of a GMPLS system for provisioning optical core and
access networks for multicast TV. In Euro“y“ yoxtorvl fiork“hop, Glasgow, UK, Mar.
2008.
Contents
Xontents xiii
aist of Figures xvii
aist of ivwles xxi
Glossvry xxiii
F Introyuxtion F
1.1 What is IPTV? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.2 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.2.1 The Value of IPTV . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.2.2 Killer application? . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2.3 Challenges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.3 Issues not covered in this thesis . . . . . . . . . . . . . . . . . . . . . . . 4
1.3.1 IPTV, not Internet TV nor P2P TV . . . . . . . . . . . . . . . . 4
1.3.2 Broadcast IPTV, not VoD . . . . . . . . . . . . . . . . . . . . . . 5
1.3.3 Single-domain IPTV, not multiple . . . . . . . . . . . . . . . . . 5
1.4 Contributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.4.1 Reducing channel change delay . . . . . . . . . . . . . . . . . . . 6
1.4.2 Reducing energy by avoiding waste . . . . . . . . . . . . . . . . . 6
1.4.3 Reducing energy by integrating optical switching . . . . . . . . . 7
1.4.4 Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.5 Outline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
G Wvxkgrouny N
2.1 Video coding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.1.1 IPTV channel change delay . . . . . . . . . . . . . . . . . . . . . 11
xiv XdciZcih
2.2 IP multicast . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.2.1 Why multicast? . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.2.2 Multicast addresses . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.2.3 Group management . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.2.4 Multicast routing protocols . . . . . . . . . . . . . . . . . . . . . 17
2.2.4.1 Flood and prune protocols . . . . . . . . . . . . . . . . 17
2.2.4.2 Centre-based trees protocols . . . . . . . . . . . . . . . 18
2.3 A typical IPTV network . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
2.3.1 Network topology . . . . . . . . . . . . . . . . . . . . . . . . . . 19
2.3.2 Network protocols . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.3.3 IPTV services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
2.4 Challenges for IPTV providers . . . . . . . . . . . . . . . . . . . . . . . 23
H htvte of the vrt GL
3.1 Measurement of IPTV systems . . . . . . . . . . . . . . . . . . . . . . . 28
3.2 A survey on techniques to reduce IPTV channel change delay . . . . . . 29
3.2.1 Simple optimisations using pure multicast . . . . . . . . . . . . . 29
3.2.2 Proxy server with boost streams . . . . . . . . . . . . . . . . . . 30
3.2.3 Video coding techniques . . . . . . . . . . . . . . . . . . . . . . . 33
3.2.4 Predictive pre-joining of TV channels . . . . . . . . . . . . . . . 34
3.3 Green networking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
3.3.1 Fundamentals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
3.3.2 No work? Then, sleep . . . . . . . . . . . . . . . . . . . . . . . . 38
3.3.3 Little work? Then, slow down: Adaptive Link Rate . . . . . . . 38
3.3.4 Want to sleep? Then, delegate: Proxying . . . . . . . . . . . . . 39
3.3.5 Change paradigm: energy-aware infrastructures . . . . . . . . . . 40
3.3.6 Shifting to save: Traffic Engineering . . . . . . . . . . . . . . . . 42
3.3.7 Lacking information? Measure it . . . . . . . . . . . . . . . . . . 43
3.4 From electronics to optics . . . . . . . . . . . . . . . . . . . . . . . . . . 44
3.4.1 Optics vs electronics: there is room for both . . . . . . . . . . . . 44
3.4.2 Hybrid architectures: the best of both worlds . . . . . . . . . . . 46
I bethoyology vny yvtvset IN
4.1 Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
4.2 Dataset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
4.2.1 Data collection . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
XdciZcih xv
4.2.2 Data characterisation . . . . . . . . . . . . . . . . . . . . . . . . 52
4.2.3 Data cleaning and parsing . . . . . . . . . . . . . . . . . . . . . . 54
4.2.4 Validation of the dataset . . . . . . . . . . . . . . . . . . . . . . . 56
5 geyuxing xhvnnel xhvnge yelvy 6F
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
5.2 Simple scheme: pre-joining neighbouring channels . . . . . . . . . . . . . 66
5.3 Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
5.4 Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
5.5 Pre-joining popular TV channels . . . . . . . . . . . . . . . . . . . . . . 72
5.6 Personalised scheme: tracking user behaviour . . . . . . . . . . . . . . . 74
5.6.1 Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
5.7 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
5.8 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
6 gesourxe vny energy effixient network MF
6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
6.2 Selective joining . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
6.3 Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
6.4 Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
6.5 Impact on energy consumption . . . . . . . . . . . . . . . . . . . . . . . 91
6.5.1 Power consumption model . . . . . . . . . . . . . . . . . . . . . . 92
6.5.2 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
6.6 Discussion: effect on channel change delay . . . . . . . . . . . . . . . . . 98
6.7 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
L dptixvl wypvss of populvr ik xhvnnels FE5
7.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
7.2 The use of optical bypass to save energy . . . . . . . . . . . . . . . . . . 109
7.3 Protocol for optical bypass in IPTV . . . . . . . . . . . . . . . . . . . . 110
7.4 Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
7.5 Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
7.5.1 Scalability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
7.5.2 Opportunities for optical bypass . . . . . . . . . . . . . . . . . . 116
7.6 Impact on energy consumption . . . . . . . . . . . . . . . . . . . . . . . 118
7.6.1 Selective joining in core optical networks . . . . . . . . . . . . . . 118
7.6.2 Energy consumption model of the hybrid nodes . . . . . . . . . . 120
xvi XdciZcih
7.6.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
7.7 Discussion: on the value of electronics . . . . . . . . . . . . . . . . . . . 123
7.8 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
M hummvry of xontriwutions vny future work FGL
8.1 Summary of contributions . . . . . . . . . . . . . . . . . . . . . . . . . . 127
8.2 Future directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
8.2.1 Improving channel change user experience . . . . . . . . . . . . . 129
8.2.2 Improving resource efficiency . . . . . . . . . . . . . . . . . . . . 130
8.2.3 Improving energy efficiency . . . . . . . . . . . . . . . . . . . . . 131
V From elextronixs to optixsO envwling texhniques FH5
A.1 Optical multicast . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
A.2 Traffic grooming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
A.3 Aggregated multicast . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
geferenxes FIF
aist of Figures
2.1 Typical frame structure . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.2 Main components of channel change time . . . . . . . . . . . . . . . . . 12
2.3 Contribution of each component of channel change time (not to scale) . 14
2.4 IGMP basic network architecture . . . . . . . . . . . . . . . . . . . . . . 16
2.5 IPTV reference architecture . . . . . . . . . . . . . . . . . . . . . . . . . 20
2.6 Panasonic’s 150” plasma TV presented at CES 2008 [210] . . . . . . . . 23
3.1 Zapping servers introduced in the IPTV network to mitigate the high
channel change delay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
3.2 Optical bypass-enabled network node . . . . . . . . . . . . . . . . . . . . 46
4.1 Number of channel switches in zapping mode . . . . . . . . . . . . . . . 50
4.2 Data collection process . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
4.3 Number of viewers during a representative week . . . . . . . . . . . . . . 57
4.4 Channel popularity distribution . . . . . . . . . . . . . . . . . . . . . . . 58
4.5 Correlation between channel access frequency and channel dwell time . . 60
5.1 Cumulative distribution of zapping jump distance . . . . . . . . . . . . . 63
5.2 Predictive pre-joining of TV channels . . . . . . . . . . . . . . . . . . . . 67
5.3 Proposed methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
5.4 Percentage of requests that experience no delay by using the simple
scheme, for various values of channel concurrent time i and number
of neighbours . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
5.5 Percentage of requests that experience no delay by using the simple
scheme during zapping periods only, for various values of channel con-
current time i and number of neighbours . . . . . . . . . . . . . . . . . 71
5.6 Performance gap between optimal predictor and the simple scheme for
various values of channel concurrent time i and number of neighbours . 72
xviii aIhi dF FIGjgZh
5.7 Average bandwidth consumed by the simple scheme for various values
of channel concurrent time i and number of neighbours, compared with
an optimal predictor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
5.8 Percentage of requests that experience no delay by pre-joining the seven
most popular channels, for a channel concurrent time i equal to two min-
utes. Different data point types are used for the hybrid (“popular chan-
nels included”) and the neighbours-only (“popular channels excluded”)
schemes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
5.9 Variance of the percentage of requests that experience no delay by using
the simple scheme for various values of channel concurrent time i and
number of neighbours. In the graph the median, 5th and 95th percentile
are presented . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
5.10 Variance of the percentage of requests that experience no delay by using
the personalised scheme for a channel concurrent time i equal to 60
seconds, 2 neighbours, and several values of . In the graph the median
and the standard deviation are presented . . . . . . . . . . . . . . . . . 77
6.1 Average number of active channels (TV channels with viewers) per net-
work node (including DSLAMs and core-regional routers). Nodes are
ordered by the number of users they serve . . . . . . . . . . . . . . . . . 84
6.2 Proposed methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
6.3 Number of channels joined when using the selective joining scheme for
various values of the inactive set size . . . . . . . . . . . . . . . . . . . . 90
6.4 Percentage of requests affected for various values of the inactive set size. 91
6.5 Power consumption model (zoomed) . . . . . . . . . . . . . . . . . . . . 94
6.6 Current router power consumption v“ energy-proportional node . . . . . 95
6.7 Power savings after introducing the proposed scheme as a function of the
baseline traffic load in the node (regional network) . . . . . . . . . . . . 96
6.8 Power savings after introducing the proposed scheme as a function of the
baseline traffic load in the node (core network) . . . . . . . . . . . . . . 97
6.9 The four scenarios considered . . . . . . . . . . . . . . . . . . . . . . . . 100
6.10 Trading relationship between the switching requests that experience no
delay and the number of channels distributed to a DSLAM. . . . . . . . 101
7.1 Core network topology considered . . . . . . . . . . . . . . . . . . . . . 108
7.2 Optical network employing optical bypass techniques . . . . . . . . . . . 109
7.3 Proposed methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
aIhi dF FIGjgZh xix
7.4 TV channel churn rate for all eleven regions, for five values of the update
interval . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
7.5 Average number of TV channels that are optically bypassed, electroni-
cally routed and not distributed, respectively . . . . . . . . . . . . . . . 118
7.6 Power savings of using the selective joining scheme considering an optical
IP network, as a function of the baseline traffic load in the node (core
network) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
7.7 Hybrid node . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
7.8 Power savings achieved by optically bypassing popular TV channels in
the network core, as a function of the baseline traffic load in the node . 123
7.9 Power savings achieved by optically bypassing all TV channels in the
network core (compared to popular TV channels only), as a function of
the baseline traffic load in the node . . . . . . . . . . . . . . . . . . . . . 124

aist of ivwles
4.1 Dataset statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
4.2 Parameters of the channel popularity models . . . . . . . . . . . . . . . 59
6.1 Description of the three scenarios . . . . . . . . . . . . . . . . . . . . . . 91
6.2 Linecard power profile . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93

Glossvry
VDha Asymmetric Digital Subscriber Line.
Vag Adaptive Link Rate.
WGe Border Gateway Protocol.
XVGg Compound Annual Growth Rate.
XWi Core Based Trees.
XXc Content Centric Networking.
XDc Content Distribution Network.
XZh Consumer Electronics Show.
Xbdh Complementary Metal Oxide Semiconductor.
DZbjm Demultiplexer.
DFh Dynamic Frequency Scaling.
Dha Digital Subscriber Line.
DhaVb DSL Access Multiplexer.
Dhb Dynamic Spectrum Management.
Dkbge Distance-Vector Multicast Routing Protocol.
Dkh Dynamic Voltage Scaling.
Ze Energy Proportional.
xxiv Glossvry
FZX Forward Error Correction.
FID First I-Frame delay.
FeGV Field Programmable Gate Array.
FiiX Fibre To The Cabinet.
FiiH Fibre To The Home.
Fiix Fibre To The x.
Flb Four Wave Mixing.
Gbeah Generalized MPLS.
Gde Group Of Pictures.
HD High Definition.
HDik High Definition TV.
IDd Input/Output.
IXi Information and Communications Technology.
IGbe Internet Group Management Protocol.
Ie Internet Protocol.
Ieik Internet Protocol Television.
Ihe Internet Service Provider.
aVc Local Area Network.
bXBglV Multicast Routing and Wavelength Assignment.
bZbh Micro Electro Mechanical Systems.
bIae Mixed Integer Linear Programming.
beah Multiprotocol Label Switching.
bjm Multiplexer.
Glossvry xxv
cvDv Nano Data Center.
cIX Network Interface Controller.
cie Network Time Protocol.
dZd Optical-Electrical-Optical.
dheF Open Shortest Path First.
dheFBiZ OSPF-Traffic Engineering.
dmX Optical Cross Connect.
eGe Peer-to-peer.
eIb Protocol Independent Multicast.
eIbBDb Protocol Independent Multicast - Dense Mode.
eIbBhb Protocol Independent Multicast - Sparse Mode.
foZ Quality of Experience.
gVb Random Access Memory.
gIe Routing Information Protocol.
ge Rendezvous Point.
ghke Resource Reservation Protocol.
ghkeBiZ RSVP-Traffic Engineering.
giXe Real-time Transport Control Protocol.
gie Real-time Transport Protocol.
glV Routing and Wavelength Assignment.
hvD Split and Delivery.
hD Standard Definition.
hDik Standard Definition TV.
xxvi Glossvry
hFXh Synchronisation Frames for Channel Switching.
heb Self Phase Modulation.
hei Shortest Path Trees.
hiW Set Top Box.
ivX Tap and Continue.
iZ Traffic Engineering.
jDe User Datagram Protocol.
jHDik Ultra High Definition TV.
kb Virtual Machine.
koD Video on Demand.
lX Wavelength Conversion.
lDb Wavelength Division Multiplexing.
oV Zapping Accelerator.
Chvpter F
Introyuxtion
In 1884, Paul Nipkow, a German engineering student, proposed and patented the
Nipkow disk, “an electric telescope for the electric reproduction of illuminating ob-
jects” [209]. Some years later, this mechanical image scanning device became the basis
for the essential component of the first television set. The importance of this invention
was emphasised by Albert Abramson, an historian of television, who considered this to
be “the master television patent” [1]. Some decades later, in 1925, John Logie Baird
gave the first public demonstration of television at Selfridges department store in Lon-
don. These two events mark the beginning of a revolution that continues today. For
more than half a century, television has been a dominant and pervasive mass media,
experimenting profound changes [42]. From mechanical to fully-electronic television,
from black-and-white to colour, from analog to digital, the technological advances have
been impressive.
The distribution of television is currently dominated by three technologies: over
the air broadcasts, cable, and satellite. The advent of IP networks and the increased
availability of broadband access created a new vehicle for the distribution of TV ser-
vices. The distribution of digital TV services over IP networks, or IPTV, offers much
more than traditional broadcast TV. The high visual quality and reliability expecta-
tions of traditional broadcast TV can now marry the interactivity, flexibility and rich
personalisation enabled by IP technologies [191]. IPTV has even been hyperbolised as
the “killer application for the next-generation Internet” [211].
The topic of this dissertation is the distribution of TV over IP networks. In the
following pages I investigate some of the challenges faced by IPTV network operators,
and I propose and analyse novel techniques to address these challenges.
G Introyuxtion
FCF lhvt is IeikT
IPTV is a method of delivering entertainment-quality video using an IP network as the
medium, instead of the hitherto predominant cable, free-to-air or satellite broadcasts.
Advances in networking technology, digital media and codecs1 have made it possible
for broadband service providers throughout the world to begin streaming live and on-
demand television to homes over their high-speed IP networks [141]. IPTV extends the
reachability of content to any IP-connected device, which today means it enables the
availability of content to almost anywhere (something users cannot get from traditional
services) [29]. As an example of this concept, several companies, such as the BBC and
Time Warner Cable Inc., launched very recently applications that allow users to watch
live TV and catch up on their favourite TV programmes on their iPads, iPhones, and
Android mobile devices [38].
FCG botivvtion
This section outlines the motivation for doing research on IPTV. In the following sec-
tions I discuss the value of IPTV and highlight the challenges faced by IPTV operators
addressed in this dissertation.
FCGCF ihe kvlue of Ieik
Until the development of IP networks, television was a broadcast medium. Traditional
TV networks offered limited freedom of choice and control to its users. Over the years,
the number of channels increased from a few free-to-air broadcasts to several hundreds,
offering a much wider selection but still effectively delivering the same service. In this
sense, IPTV reinvents television [191]. Its integral return channel and its ability to
address individual users paves the way for new interactive services.
This bidirectional communication capability also gives more visibility on viewing
activities, allowing the service provider to know what the users are watching and when.
This raises several issues, such as privacy, but can be a catalyst for the creation of
new applications. Television advertisement, for instance, can be reinvented [191]. For
telecommunications operators, IPTV offers flexibility and added value in the form of ad-
ditional services that can be offered to its customers, which improves their profitability
and competitive edge [141].
1V xoyzx i“ v yzvixz xvpvwlz of znxoying or yzxoying v yigitvl mzyiv “trzvmC
FCG botivvtion H
FCGCG Killer vpplixvtionT
The past few years have witnessed the rapid roll-out of IPTV services. IPTV has been
launched by major service providers worldwide — France Telecom, AT&T, Telefonica,
China Telecom, Korea Telecom, among others [43] — and its popularity is on the
rise [201]. In the United States, for instance, there are already more than 5 million
subscribers, and this number is expected to increase to 15.5 million by 2013 [156]. By
early 2009, there were more than 25 million IPTV users in the world [138]. IPTV
provider managed traffic is expected to grow at a Compound Annual Growth Rate
(CAGR) of 53% for the next few years [56].
Contrary to other industries (newspapers, music industry, book publishers) TV is
coping well with technological change [71]. An average TV viewer spends 5 hours per
day in front of the box, five times more than using the Internet [148]. On February
17th 2010, 106 million Americans watched the Super Bowl — a record for a single
program. Tokyo residents spend more time consuming media online (from 6 minutes
in 2000 to one hour in 2009), but the time spent in front of the TV is also growing —
now to an average of 216 minutes [71]. Television is therefore still supreme at holding
the attention of a large number of people for long periods.
In 1996, George Gilder, an American writer, claimed that by the end of the twentieth
century television would be extinct due to technological advances. From his book, “Life
after Television” [86]:
qVll thz“z yzvzlopmznt“ xonvzrgz in onz kzy fvxt of lifzA vny yzvthA for tzlzxommuB
nixvtion“ in thz FNNE“C izlzvi“ion vny tzlzphonz “y“tzm“ — optimizzy for v fiorly in
fihixh “pzxtrum or wvnyfiiyth fiv“ “xvrxz — vrz uttzrly un“uitzy for v fiorly in fihixh
wvnyfiiyth i“ vwunyvntC7
The facts seem to contradict Gilder’s assertion. As with Mark Twain’s reports on
his own death, Gilder’s claims on the death of television seem to be exaggerated.
FCGCH Chvllenges
IPTV distribution imposes stringent requirements on both performance and reliabil-
ity, requiring low latency, a tight control of jitter, and small packet loss in order to
guarantee the expected video quality. The offer of this service is therefore challenging
for IPTV operators that want to match the level of quality of service that customers
are accustomed to from other TV service providers. The problem is that IP networks
are “best effort”, susceptible to lost or dropped packets as bandwidth becomes scarce
and jitter increases. This challenge is partially solved by current IPTV networks be-
I Introyuxtion
ing provider-managed services. In their “walled garden” IPTV infrastructures service
providers control the load of the network elements and use traffic prioritisation and
bandwidth reservation techniques to assure service quality and performance. But other
problems exist in this respect. A major concern of IPTV operators is channel change
delay. This is the latency experienced by users when switching between channels. Due
to bandwidth limitations, in current IPTV networks only one or two TV channels are
distributed in the access link that connects the network to the Set Top Box1(STB).
When a user switches to a new channel, the STB has to issue a new channel request
towards the network. This is one of the causes of channel change delay. In addition to
this network delay, synchronisation and buffering of media streams can cause channel
change delays of several seconds. This is the first challenge I address in this dissertation,
in Chapter 5.
Video distribution is very resource intensive. High definition TV requires bit rates
on the tens of Mbps range, and future ultra high definition formats may increase this
figure by orders of magnitude. Efficiency in distribution is therefore a major concern.
The emergence of scalable multicast protocols has provided the means for an efficient
distribution of TV services. However, current IPTV multicast architectures remain
inefficient. They use “tvtix multicast, distributing vll TV channels from the source to
every access node in the network continuously. As particular channels have no viewers
at particular time periods, this method is provably rz“ourxz and znzrgy inefficient.
These inefficiencies are the second challenge for IPTV network providers I address in
this dissertation, in Chapters 6 and 7.
FCH Issues not xoverey in this thesis
Research on IPTV covers a broad range of interesting topics. The term IPTV itself has
been used in the research community to mean very different things. To try to clarify
its precise topic, I provide in this section an explanation of what this dissertation i“
not.
FCHCF IeikA not Internet ik nor eGe ik
The term IPTV is sometimes used in certain contexts to describe WebTV, Internet
TV or P2P-based TV. Internet TV and WebTV are normally used to describe the
delivery of TV programming over the public Internet, typically to personal computers as
1ihz yzvixz thvt turn“ thz pvxkzt“ rzxzivzy from thz nztfiork into xontznt fihixh i“ thzn yi“plvyzy
on thz tzlzvi“ion “xrzznC
FCI Xontriwutions 5
streamed or downloadable video content [5]. Broadcasters such as the BBC are already
providing this type of service over the Internet [25]. This approach is also referred
to as over-the-top (OTT) video, since it essentially uses the Internet as a transport
pipe to deliver content. As opposed to PC-based viewing, IPTV services target a TV
viewing environment integrated with set-top boxes (STBs), providing cable TV-like
experience. The distribution of these services is done in closed, privately-managed
IPTV networks. The service provider has full control over content distribution, storage
management, and bandwidth provisioning, to ensure end-to-end quality of delivery. In
addition, IPTV is a server-centred architecture, in contrast with P2P-based TV. In this
dissertation, I target IPTV, not Internet TV nor P2P TV.
FCHCG Wrovyxvst IeikA not koD
IPTV services are usually classified into two main types: broadcast television and
Video-on-Demand (VoD). The VoD service model is one-to-one. A single copy of a spe-
cific program is unicast to a single subscriber on request. In contrast, IPTV broadcast
services require all viewers to watch a program simultaneously, according to a prede-
termined schedule. This is a one-to-many service model where for efficiency reasons
IP multicast is used for distribution. In this dissertation, I target broadcast television
services, not VoD.
FCHCH hingleByomvin IeikA not multiple
Routing in the Internet forms a two level hierarchy: inter and intra-domain. Traffic
crossing multiple network domains is governed by policy-based border gateway protocol
(BGP) [168]. As I referred before, due to its stringent quality of service requirements,
IPTV is currently a service managed by a single provider. The design techniques
proposed in this dissertation apply therefore to a single independently operated network
domain.
FCI Contriwutions
In this dissertation I propose and analyse several techniques to assist IPTV providers
in the design of novel resource and energy efficient networks. These techniques focus
on the technological challenges referred to before: IPTV service’s high channel change
delay and network efficiency. The main contributions of this dissertation are as follows.
6 Introyuxtion
FCICF geyuxing xhvnnel xhvnge yelvy
The first contribution of this dissertation is an empirical analysis of a particular solution
to the channel change delay problem, namely, predictive pre-joining of TV channels.
In this scheme each Set Top Box simultaneously joins additional multicast groups (TV
channels) along with the one requested by the user. If the user switches to any of
these channels next, switching latency is virtually eliminated, and user experience is
improved. Previous work on this subject used simple mathematical models to perform
analytical studies or to generate synthetic data traces to evaluate these pre-joining
methods. By analysing IPTV channel switching logs from an IPTV service offered by
an operational backbone provider, I demonstrate that these models are conservative
in terms of the number of channel switches a user performs during zapping periods.
They therefore do not evidence the true potential of predictive pre-joining solutions.
To fill this gap I perform a trace-driven analysis using the dataset referred to above
(the switching logs) to evaluate the potential of these solutions. The main conclusion of
this study is that a simple scheme where the neighbouring channels (i.e., the channels
adjacent to the requested one) are pre-joined by the Set Top Box alongside the requested
channel, during zapping periods only, eliminates zapping delay for around half of all
channel switching requests to the network. Importantly, this result is achieved with a
negligible increase of bandwidth utilisation in the access link [163, 164].
FCICG geyuxing energy wy vvoiying wvste
Current IPTV service providers build “tvtix multicast trees for the distribution of TV
channels. By static multicast I mean that all receivers are known beforehand, and no
new group members are allowed to join - it is a “tvtix set of receivers for all TV content.
This means all TV channels are distributed everywhere in the network continuously.
This is justified to guarantee the quality of experience required by IPTV customers.
By distributing TV channels to as close to the users as possible, network latencies do
not add significantly to the already high channel change delay. However, as particular
channels have no viewers at particular time periods, this method is provably resource
and energy inefficient. To reduce these inefficiencies, I propose a semi-dynamic scheme
where only a selection of TV multicast groups is distributed in the network, instead
of all. This selection changes with user activity. This method is evaluated empirically
by analysing the same dataset as above. I demonstrate that by using the proposed
scheme IPTV service providers can save a considerable amount of bandwidth while
affecting only a very small number of TV channel switching requests. Furthermore, I
FCI Xontriwutions L
show that although today the bandwidth savings would have reduced impact in energy
consumption, with the introduction of numerous very high definition channels this
impact will become significant [165].
FCICH geyuxing energy wy integrvting optixvl switxhing
The third contribution of this dissertation is a novel energy friendly protocol for core op-
tical IPTV networks. The objective is to further increase the energy efficiency of IPTV
networks. The fundamental concept is to blend electronic routing and optical switching,
thus gluing the low-power consumption advantage of circuit-switched all-optical nodes
with the superior bandwidth-efficiency of packet-switched IP networks. The main idea
is to optically switch popular TV channels. These channels are watched by many, hav-
ing viewers everywhere in the network at any time. These are long-lived flows in the
network, and are therefore perfect targets for this type of slow energy-friendly switch-
ing. With the use of this protocol, popular IPTV traffic optically bypasses the network
nodes, i.e., this traffic avoids electronic processing. I evaluate this proposal empirically
by performing a trace-driven analysis using the IPTV dataset mentioned before. The
main conclusion is that the introduction of optical switching techniques results in a
quite significant increase in the energy efficiency of IPTV networks.
FCICI Evvluvtion
All the schemes presented in this dissertation are evaluated by means of trace-driven
analyses using a dataset from an operational IPTV service provider, Telefonica. This
dataset was obtained from measurements collected by Telefonica in its network, from
April 2007 to October 2007. The traces recorded user channel change activity from
Telefonica’s IPTV service, Imagenio. The dataset scales up to 150 TV channels, six
months, and 255 thousand users. It is widely accepted that a thorough evaluation
using real workloads enables the assessment of future network architectures with an
increased level of confidence. This is particularly relevant in a research field that has
relied heavily upon hypothetical user models which are different from the reality and
can lead to incorrect estimation of system performance. I believe a strength of this
dissertation lies in such thorough evaluation using real traces of real IPTV usage.
M Introyuxtion
FC5 dutline
This dissertation is organised as follows. In Chapter 2, I present some background to
the distribution of TV in IP networks. Then, in Chapter 3, I describe relevant research
related to IPTV with a particular focus on the challenges addressed and the techniques
proposed in this dissertation. In Chapter 4, I justify the option for the methodology
used and I describe the dataset used for evaluation. In particular, I detail the process
of data collection, data cleaning, and how the dataset is validated. Next, I address
the problem of channel change delay in IPTV networks, in Chapter 5. In particular,
I present an in-depth analysis of predictive pre-joining solutions to this problem. In
Chapter 6, I propose a semi-dynamic multicast scheme to increase IPTV network’s
resource and energy efficiency. To increase energy efficiency further, in Chapter 7, I
assess the opportunities for introducing optical bypass in core optical IPTV networks
and demonstrate its effectiveness in reducing energy consumption. Finally, in Chapter
8, I summarise the contributions of this work and discuss possible directions for future
research.
Chvpter G
Wvxkgrouny
The topic of this dissertation is the distribution of TV in IP networks. There are two
aspects to this issue: the content and the delivery. In the first section of this chapter, I
address the former, explaining why video xontznt needs to be coded and how it is coded.
I also explain its implications in increasing channel change delay in IPTV networks.
Then I cover the yzlivzry of TV services using IP multicast. I describe the multicast
service model and the most common multicast protocols. Next, I describe a typical
IPTV architecture in some detail. Finally, I conclude the chapter with a summary of
the challenges faced by IPTV providers which were the motivation for this work.
GCF kiyeo xoying
Video consists of a series of pictures, or frames, taken at regular intervals (typically
every 33.3ms or 40ms [65]). The data rate of this raw signal is too high for econom-
ical transport over telecommunication networks. Uncompressed Standard Definition
TeleVision (SDTV), for instance, requires a bit rate of around 200 Mbps, and High
Definition TeleVision (HDTV) already demands bit rates close to 1 Gbps [83].
Since network bandwidth is a scarce resource, compression techniques are needed to
save transmission capacity (and storage). Efficient media coding schemes have therefore
been developed, such as MPEG-2 [109] and MPEG-4 [110]. They make use of the
fortunate fact that much audio and video is redundant, containing, effectively, repeated
or less useful data, which leads to a high correlation between adjacent frames in a typical
video sequence. Hence, dependencies between neighbouring frames can be exploited to
increase coding efficiency.
In these coding schemes the video streams are divided into segments, each com-
FE Wvxkgrouny
monly termed a Group of Pictures (GOP), as in Figure 2.1. The video is coded, with
three types of frames defined: I, P and B-frames. A GOP is composed of all the pre-
dicted frames (P and B) between two I-frames, together with the starting I-frame. The
GOP size is thus defined as the time between I-frames. Each type of frame explores a
different redundancy pattern existing in video sequences and therefore results in differ-
ent compression efficiencies and in different functionality. I-frames1 are encoded with
image compression techniques that exploit the spatial correlation of pixels within the
frame without using information from any other frames. Since they are not dependent
on any other frame, they are used as a decoding reference for other frames, and can
serve as access points where decoding can begin. P and B-frames are predicted based
on one or more surrounding frames, using the estimated motion of objects of the frame
it refers to. Therefore, they cannot be decoded alone. P-frames use motion prediction
from a past reference frame and B-frames use a prediction based on references from the
past, future, or a combination of both.
I-frame
B-frame
P-frame
GOP
Figure 2.1: Typical frame structure
To decode a video stream the decoder will therefore need an I-frame as a first
reference frame, which can be decoded without further information. To speed up play
out time, it would be advantageous to transmit I-frames very frequently. This way
decoding and play out could start sooner, reducing inconvenient delays. The problem
is that I-frames are significantly larger than P or B-frames, requiring higher storage
space and higher bit rates to be transported. Depending on the content, this difference
can be of one order of magnitude [182]. There is thus a trade-off between compression
efficiency on one side and play out performance on the other. In practise, GOP duration
1ihz“z frvmz“ vrz vl“o xvllzy vnxhorBfrvmz“ or kzyBfrvmz“A vlwzit in “omztimz“ yiffzrznt xontzxt“C
[or in“tvnxzA vn IBfrvmz i“ vlfivy“ v kzyBfrvmz in beEGBGA wut thi“ i“ not v “uffixiznt xonyition in
beEGBIC In thz follofiingA hofizvzrA I fiill xon“iyzr thz beEGBG “tvnyvry vny u“z only thz tzrm
IBfrvmzC
GCF kiyeo xoying FF
is typically in the range of 1 to 2 seconds [65, 182]. More advanced codecs, however,
have longer GOPs to gain from the encoding efficiency, at the cost of higher latencies.
GCFCF Ieik xhvnnel xhvnge yelvy
In traditional analogue TV broadcast and cable technology, channel change is almost
instantaneous since it only involves the TV receiver tuning to a specific carrier fre-
quency, demodulating the content and displaying it on the TV screen. The zapping
delay in these systems is typically less than 200ms [28]. TV viewers thus consider
zapping times to be virtually instant, and have become used to this surfing (through
channels) experience. With the digitisation and compression of content, zapping times
have increased significantly. Users already experience this today in digital cable TV
networks, but it is an even more severe problem in IPTV, since zapping times are
also affected by network delay. Users zapping in IPTV usually experience a couple of
seconds’ delay or more [176].
Kooij zt vlC [128] presented a study recently where they conclude that to achieve an
acceptable quality of service, channel change time needs to be below 0.43 s. Although
the study is limited in terms of the size of the test subject population, it is clear that
IPTV high zapping delay degrades the quality of experience perceived by customers and
is a major obstacle for IPTV services wide adoption. To understand how to mitigate
this problem, it is important to understand the components of channel change time.
Consider Figure 2.2, where the main contributors to zapping delay are depicted.
The user starts by issuing a channel change request using the remote control. The
request reaches the Set Top Box (STB) after an estimate 5-10ms delay [199] (step 1 in
the figure).
In IPTV video delivery, in contrast to cable networks, for instance, typically only
the channel the user is watching is delivered to the STB at any one time. This is due to
bandwidth limitations in the access network. When a user switches to a new channel,
the STB has to issue a new channel request towards the network (step 2). Since video
distribution is done via multicasting, this is translated into leave and join multicast
requests. In typical systems, these operations are handled by a group management
protocol, usually IGMP (Internet Group Management Protocol) [39]. This protocol is
analysed later in this chapter. In short, the STB sends a leave request from the current
multicast session (the TV channel the user is currently watching) and a join request to
the new multicast group (the TV channel the user is switching to). This channel change
request reaches the first upstream network node that has the channel available and the
FG Wvxkgrouny
1. Channel change
request
2. IGMP Leave/Join
3. Synchronisation delay
4. Video buffer delay
5. STB processing delay
IPTV
network
Set Top BoxRemote control
Figure 2.2: Main components of channel change time
routing infrastructure sets up the multicast forwarding state to deliver the packets of
the multicast session to the STB. To minimise this component of the delay, vll TV
channels are distributed very close to the user, commonly to the DSLAM1 (in DSL
networks2) or to the local router. In addition to this “ignvlling yzlvy we need to add
the propvgvtion yzlvy experienced in the access link. The sum of these types of delay,
which I jointly call nztfiork yzlvy, is usually below 100-200ms [27, 28, 80, 182, 189].
This is therefore a relatively unimportant contributor to the overall delay, as is made
clear in the following paragraphs.
After the STB receives the first packets from the recently joined multicast group,
there is still a time lag before it can start consuming the audiovisual data because
the STB must wait for the next I-frame before it can start decoding the content, as
explained in the previous section. I refer to this as the “ynxhroni“vtion yzlvy (step
3). The maximum synchronisation delay is equal to the duration of the GOP, which
occurs when the STB just misses the start of an I-frame and thus has to wait for the
next. On average, this delay is half the GOP duration. Synchronisation delay therefore
comprises a substantial portion of the channel change time (recall that GOP duration
1ihz Dha Vxxz““ bultiplzxzr i“ v lvyzrBG vggrzgvtion “fiitxh thvt xonnzxt“ multiplz xu“tomzr Dha
intzrfvxz“ to thz nztfiorkC
2In thi“ typz of vxxz““ nztfiork thz yigitvl yvtv i“ trvn“mittzy ovzr tfii“tzyBpvir xoppzr fiirz“ of v
loxvl tzlzphonz nztfiorkA u“ing “zpvrvtz frzquznxy wvny“ from thz tzlzphonz “ignvl“C
GCF kiyeo xoying FH
is typically in the range of 1 to 2 seconds). Many video services also employ content
encryption, so the encryption keys must be acquired and provided to the decryption
engine for decrypting the content, and this also adds to synchronisation delay.
In general, multicast-based video applications use an unreliable underlying trans-
port protocol such as UDP [158] to distribute IPTV content. For this reason, packet
loss may occur and loss-repair techniques need to be included in the system. For ex-
ample, a local repair server can be included at the network edge for retransmitting
lost packets or, if the retransmission cost is high, Forward Error Correction (FEC)
techniques1 may be used to provide reliability. Regardless of the type of loss-repair
method, buffering will be required at the receiver side for these operations to be per-
formed (step 4). Buffering reduces the system sensitivity to short term fluctuations in
the data arrival rate by absorbing variations in end-to-end delay and allowing margins
for retransmissions when packets are lost.
There are three other reasons to buffer incoming packets before forwarding them to
the decoder: to avoid under-run (starvation) and to compensate for network jitter and
packet-reordering delay. Starvation results from the different frames being encoded at
different data rates, and thus the video encoding process resulting in a variable bit rate
stream. Since the network flow is typically constant bitrate (or capped variable), this
mismatch between the input and the output of the encoder is solved with the inclusion
of a smoothing buffer. Network jitter and packet-reordering are caused by cross-traffic
in network equipment: in an IP network traffic is asynchronous, so packets have to
wait in buffers. Also, different packets may follow different paths, and this results in
a variable and unpredictable delay between packets. While the amount of protection
offered by a buffer grows with its size, so does the latency it introduces. Typical decoder
buffer requirements range from 1 to 2 seconds [28].
The final source of delay is end-system delay (step 5). This is the processing delay
in the STB and display device. This can occur at a number of system layers, and is
generally a trade off between terminal resources (memory and processor speed) and
cost. The processing time depends very much on the STB, but 150ms is a typical
figure [182].
Figure 2.3 pictorially summarises the contribution of each component of channel
change time (not to scale). As can be observed, the main contributors to IPTV chan-
nel change latency are stream synchronisation and buffering (steps 3 and 4 in Figure
2.2), adding up to around 2 seconds on average [80, 182, 189]. The main concern in
1lith [EX thz “znyzr vyy“ rzyunyvnt yvtv to it“ mz““vgz“ vllofiing thz rzxzivzr to yztzxt vny
xorrzxt zrror“A vt thz xo“t of highzr wvnyfiiyth rzquirzmznt“C
FI Wvxkgrouny
the industry and in the research community has been, in fact, to try to improve the
performance on these two aspects [30]. Chapter 5 of this dissertation is also devoted
to this problem.
Channel 
change 
request 
IGMP 
leave/join 
Synchronisation delay 
STB 
processing 
delay 
Video buffer delay 
5-10 ms 100-200 ms 500-1000 ms 1000-2000 ms  !"#$ms 
Figure 2.3: Contribution of each component of channel change time (not to scale)
GCG Ie multixvst
This section covers the second aspect concerning the distribution of TV services in an
IP network: the delivery. I briefly explain IP multicast and some of the most important
multicast protocols developed over the years. In particular, I focus on the limitations of
the original multicast protocols and how a new class of protocols enabled the emergence
of large scale IPTV networks.
GCGCF lhy multixvstT
TV broadcasting requires all viewers watch a program simultaneously, according to a
predetermined schedule. In an IP network, the simplest method to send data to many
receivers simultaneously is to send them multiple times from the source. This method
has, however, several drawbacks. First, it is very expensive to the sender. Second,
it is very inefficient as an excessive number of duplicate packets can be carried in the
network links. Third, sending multiple unicasts requires the sender to know the address
of each and every single receiver. For all these reasons, this simple technique does not
scale.
Instead of sender replication, a better option is for the responsibility of replication to
move to the network. The simplest such scheme is broadcast: network nodes replicate
all broadcast packets (with some restrictions to avoid loops), thus ensuring packets
are delivered to all devices on the network. This solution removes the burden from
the sources but it is still very inefficient and does not scale. All nodes in the network
receive the data, including those not interested.
IP networks offer an alternative solution: multicast. This mechanism provides an
efficient many-to-many distribution of data making it the ideal solution to distribute
TV services. The original work on IP multicast routing was by Steve Deering [62]. In
GCG Ie multixvst F5
his PhD thesis [64] he presented a new service model for multicast (which he called the
Host Group Model) and a set of multicast routing algorithms to support that service
model. Since then, the multicast problem has been extensively studied and several
protocols proposed. The IP multicast model can be summarised in the following three
points [58]:
• Senders send to a multicast address.
• Receivers express interest in a multicast group address.
• Routers conspire to deliver traffic from the senders to the receivers and optimise
(for some definition of “optimise”) packet replication.
In the next subsections I address each of these.
GCGCG bultixvst vyyresses
IP unicast packets are transmitted with a source and destination address, which enables
routers to find a path from sender to receiver. To send multicast traffic the destination
address has different semantics to a unicast address: it does not represent a particular
destination. Instead, it is a group address (i.e., a logical address) that represents the
set of receivers.
GCGCH Group mvnvgement
To receive multicast traffic an interested host has to inform its local router of its in-
terest. This is done by means of a group management protocol, typically the Internet
Group Management Protocol (IGMP). When a host wants to join a multicast group it
programs its Ethernet interface to accept the relevant traffic, and sends an IGMP join
message on its local network. This informs any local router that there is a receiver for
this group now on this subnet. The local routers then arrange for the traffic destined
to this address to be delivered on the subnet.
The routers periodically send an IGMP query to this multicast group to understand
if there are still hosts interested in receiving multicast traffic from this group. If the host
is still a member, it replies with a join message (unless any other host in the subnet
does it first). In its original version [62], when a host wanted to leave a multicast
group it would need to reprogramme its Ethernet interface to reject the traffic, but
packets would still be sent to the subnet until the next IGMP query was sent by the
local router (for which no-one would respond). Joining a multicast group was therefore
F6 Wvxkgrouny
quick, but leaving could be slow. IGMPv2 [74] improves over IGMPv1 by adding the
ability for a host to signal desire to leave a multicast group, by means of an explicit
leave message. This also avoids the need for the local router to send the periodic IGMP
queries referred to above. IGMPv3 [39, 103] improves over IGMPv2 mainly by adding
the ability to listen to multicast originating from a specific set of source IP addresses
only. A network designed to deliver an IPTV multicast service using IGMP typically
uses the basic architecture presented in Figure 2.4.
IPTV video server
L2
switch
First
router
Local
router
STB
IP network
IGMP IGMPPIM-SM
IPTV multicast traffic
(all TV channels)
(a single TV
channel)
Figure 2.4: IGMP basic network architecture
It is important to note that IGMP is utilised between the client computer and a local
multicast router. A multicast routing protocol, typically PIM-SM as I explain later,
is then used in the IP network to direct multicast traffic from the IPTV server to its
multicast clients (the STBs). A final detail also worth mentioning is the IGMP snooping
capability some layer-2 switches possess. As its name implies, IGMP snooping [53] is
the process of listening to IGMP network traffic. By using this technique the switch can
maintain a map of which links need which IP multicast streams, meaning traffic can
be filtered. This prevents hosts on a local network from receiving traffic for a multicast
group they have not explicitly joined, all at layer 2.
This technique is especially useful for bandwidth-intensive IP multicast applications
such as IPTV. Current IPTV systems distribute vll TV channels to vll local routers.
All this traffic reaches a layer-2 aggregation switch (a DSLAM1 in DSL networks) for
channels to be distributed to as close to the user as possible, in order to reduce channel
change latency. If this switch has IGMP snooping filtering capabilities, as is usually
the case, it is thus possible to distribute to the Set Top Box only the TV channel the
user has switched to, filtering all others. This allows the IPTV system to overcome the
1Dha Vxxz““ bultiplzxzrC
GCG Ie multixvst FL
bandwidth limitations of access networks1.
GCGCI bultixvst routing protoxols
For multicast traffic to be delivered the routers have to build distribution trees from the
senders to all receivers of each multicast group. As the senders do not know who the
receivers are (they just send their data) and the receivers do not know who the senders
are (they just ask for the multicast traffic) the routers have to build these trees without
help from the hosts. Two generic solutions have been proposed to this problem. In the
first, flooy vny prunz, the senders flood their data to all possible receivers and have
the routers for networks where there are no receivers to prune off their branches from
the tree. In the second, xzntrzBwv“zy trzz“, explicit distribution trees are built centred
around a particular router.
GCGCICF Flooy vny prune protoxols
In these protocols the sender floods traffic throughout the network. A router may
receive the same traffic in different interfaces, rejecting any packet that arrives at any
interface other than the one it would use to send a unicast packet back to the source,
a technique known as gzvzr“z evth Forfivrying. The router then sends a copy of each
non-rejected packet out of each interface other than the one back to the source. In
this way the data are received by all routers in the network. This includes those that
have no hosts interested in receiving this traffic. As those routers know they have no
receivers (via IGMP) they then send prune messages back towards the source to stop
unnecessary traffic from flowing. The final distribution tree is what would be formed by
the union of shortest paths from each receiver to the sender, i.e., a reverse shortest-path
tree.
Two well-known protocols fall in this category: the Distance-Vector Multicast Rout-
ing Protocol (DVMRP) [203], a multicast extension to the Routing Internet Protocol
(RIP) [139]; and the Protocol Independent Multicast, Dense Mode (PIM-DM) [2]. The
main difference between these two protocols is that DVMRP computes its own unicast
routing table while PIM-DM uses that of the underlying unicast routing protocol (the
reason for being called inyzpznyznt).
Multicast protocols based on the flood and prune technique build efficient trees, but
have problems. Sending traffic everywhere and requiring routers not on the delivery
tree to store prune state is not a scalable mechanism. But for groups where most
1Vn v“pzxt yi“xu““zy lvtzr in thi“ xhvptzrC
FM Wvxkgrouny
routers actually do have receivers (where receivers are yzn“zly distributed), this type
of protocol is a good option.
GCGCICG XentreBwvsey trees protoxols
Rather than flooding the data everywhere, algorithms in the centre-based tree category
map the multicast group address to a particular unicast address of a router. Then,
explicit distribution trees centred in this router are built.
The earliest such protocol was Core-Based Trees (CBT) [20, 21], which works as
follows. To join a multicast group a CBT router sends a join message towards the
core router for the group. At each router on the way to the core, forwarding state
is instantiated for the group and an acknowledgement is sent back to the previous
router. This procedure builds the multicast tree. CBT builds wiyirzxtionvl “hvrzy
trees. Routing state is wiyirzxtionvl as packets can flow both up the tree towards the
core or down the tree away from the core, depending on the location of the source. In
addition, the tree is “hvrzy by all sources of the group.
The main advantage of CBT is the state routers need to keep. Only routers in the
distribution tree for a group keep forwarding state for that group. This protocol is
therefore highly scalable, and is especially suited for sparse groups where only a small
proportion of subnets have members. The main problem of CBT is core placement.
Without good core placement the trees constructed can be quite inefficient.
After CBT, several related protocols were proposed that took advantage of the
good scalability offered by centre-based protocols while simultaneously trying to avoid
the dependency of the core and reduce the efficiency concerns associated. The most
successful was undoubtedly [178] the Protocol Independent Multicast - Sparse Mode
(PIM-SM) [73]. The important insight of this protocol was to realise that the problem
of discovering the senders could be separated from building efficient trees.
In a similar manner to CBT, in PIM-SM when a receiver joins a group its local router
sends a join message to the core (in PIM-SM, the core router is called the Rendezvous
Point, or RP), instantiating forwarding state for the group. Contrary to CBT, however,
this state is unidirectional. It can only be used by packets flowing from the RP towards
the receiver. When a sender starts sending data it encapsulates each packet in another
IP packet and unicasts it directly to the RP. These data are de-encapsulated and then
flow down the shared tree to all receivers.
As in CBT, these unidirectional trees may not be good distribution trees, but at
least serve the purpose of starting data flowing from the senders to the receivers. Once
GCH V typixvl Ieik network FN
these data are flowing, a receiver’s local router can initiate a transfer from the shared
tree to a shortest-path tree by sending a source-specific join message towards the source
(as the receiver now knows who the source is after receiving its data). When data
starts to arrive along the shortest-path tree, a prune message is sent back up the
shared tree to avoid receiving redundant traffic. The trigger to move from the shared
tree to the shortest-path tree is adjustable, allowing a good compromise between tree
efficiency and router state scalability. For example, it may be preferable to switch high-
bandwidth multicast traffic to the shortest-path tree, as efficiency is very important in
such scenario. For low-bandwidth traffic tree efficiency is less relevant and thus reducing
router state with a shared tree may be preferable. Because PIM-SM can optimise its
distribution tree in such way it is less critically dependent on core location.
GCH V typixvl Ieik network
Unlike Internet video which runs on top of the best-effort Internet, IPTV is a provider-
managed service with strict quality-of-service requirements [29]. The service provider
has full control over content distribution, storage management, and bandwidth provi-
sioning, to ensure end-to-end quality of delivery. Incumbent operator’s IPTV networks
are therefore “walled gardens”, well provisioned to guarantee the user experience re-
quired by TV viewers [42]. In this section I present the architecture of a typical IPTV
network.
GCHCF cetwork topology
A traditional “walled garden” IPTV network can be split logically into three main
domains — the access network, the metropolitan network, and the IP network. The
IP network usually has a two-level, hierarchical structure [79]: the regional network
(sometimes called the edge or gateway) and the core (or backbone). This is shown in
simplified form in Figure 2.51.
In an IPTV system, live TV streams are encoded in a series of IP packets and de-
livered through an IP network to the residential broadband access network. The IPTV
head-end, the primary source of television content, digitally encodes video streams
received externally (e.g., via satellite) and transmits them through a high-speed IP
network. The core network comprises a small number of large routers in major pop-
ulation centres. The core routers of any one network are often highly meshed, with
1ihi“ figurz fiill wz u“zy v“ thz IPTT pcdcpclac apafgrcarspc throughout thi“ yi““zrtvtionC
GE Wvxkgrouny
core IP network
IPTV
head-end
core
network
regional
network
metro
network
TV
channels
STB
access
network
router
DSLAM
Figure 2.5: IPTV reference architecture
high-capacity WDM fibre links interconnecting them. The topology of the core typ-
ically consists of a set of nodes connected by high bandwidth 10 Gbps and 40 Gbps
links [14]. In the regional network routers are normally lower-end routers with high
port density, where IP customers get attached to the network. These routers aggregate
the customer traffic and forward it toward the core routers [79].
The metro network serves as the interface between the regional network and the
access network. Metro (typically Ethernet) switches concentrate traffic from a large
number of access nodes and uplink to two or more regional routers (to provide redun-
dancy). The access network connects each home to one of the edge switches in the
provider’s network. There is a wide variety of access technologies: from ADSL (Asym-
GCH V typixvl Ieik network GF
metric Digital Subscriber Line1) to fibre-based solutions (FTTx2) to wireless options.
The bandwidth of each access link is limited, and it varies with the technology: around
20 Mbps for ADSL3, but increasing to the hundreds of Mbps as optics comes closer
to the home. IPTV is currently being rolled out predominantly by incumbent oper-
ators [201], so ADSL has been the main access network used to distribute content to
customer premises. Since IPTV is naturally agnostic to the layers below IP, IPTV
deployments from other providers are expected in the future [191]. In ADSL, the cop-
per pairs originally installed to deliver a fixed-line telephone service are now used to
also deliver a broadband service [45]. These copper pair-based access technologies are
limited in capacity by usable bandwidth and reach, so typically only the TV channel
the user is watching is delivered to the STB (Set Top Box) at any one time. This is the
main reason why channel zapping delay is high, as is explained in Chapter 5. For these
access technologies, the terminal unit (DSLAM in ADSL networks) commonly takes
the form of a layer 2 switch with IGMP snooping capability, as mentioned before, with
line cards appropriate to the access technology facing the subscriber.
Finally, inside a household, a residential gateway connects to a modem and one or
more STBs, receiving and forwarding all data, including live TV streams, STB control
traffic, VoIP and Internet data traffic. Finally, each STB connects to a TV.
In this dissertation I use the terminology shown in Figure 2.5 to determine events at
different aggregation levels. Namely, a DSLAM serves multiple STBs, a regional-metro
router serves multiple DSLAMs, a regional-core router serves multiple regional routers,
and, lastly, the IPTV head-end serves content to all core routers.
GCHCG cetwork protoxols
As explained in the previous section, in IPTV systems the TV head-end injects live
TV streams encoded as IP packets to the IP network core. The TV channels are
distributed from the TV head-end to edge nodes (DSLAMs in Figure 2.5) through
bandwidth-provisioned multicast trees, for efficient distribution. By far, the most com-
mon multicast routing protocol [178] is PIM-SM [73]. Current networks use “tvtix IP
multicast within a single network domain. By static multicast I mean all receivers
1VDha i“ v typz of Dha tzxhnology thvt offzr“ highzr wit rvtz“ tofivry thz xu“tomzr przmi“z“
=yofin“trzvm) thvn thz rzvzr“z =up“trzvm)C
2ihi“ i“ v gznzrix tzrm for vny wrovywvny nztfiork vrxhitzxturz u“ing optixvl fiwrz to rzplvxz vll or
pvrt of thz u“uvl xoppzr loxvl loop u“zy for lv“t milz tzlzxommunixvtion“C Exvmplz“ inxluyz“ [ii]
=[iwrz io ihz ]omz) vny [iiX =[iwrz to thz Xvwinzt)C
3V figurz thvt vvriz“ fiith thz quvlity of thz tfii“tzyBpvir loxvl loop vny it“ lzngth =iCzCA thz yi“tvnxz
from thz hou“zholy to thz loxvl zxxhvngz)C
GG Wvxkgrouny
are known beforehand, and no new group members are allowed to join — we have a
“tvtix “zt of rzxzivzr“ for all TV content. Again referring to Figure 2.5, this means vll
DSLAMs join vll multicast groups (thus receive content from vll TV channels). This is
despite the fact that particular channels may have no viewers at particular time periods.
The only section of the network which is not static in this sense is between the DSLAM
(or local router, in case the layer 2 switch has no snooping capabilities) and the STB,
due to the limited bandwidth resources of access networks referred above. Therefore,
when a user switches to a new channel, the STB issues a new channel request towards
the network.
Distributing unwanted traffic (the TV channels for which there are no viewers)
in the network may seem strange as it represents an inefficient use of the network’s
resources, with plausible energetic and monetary costs. But there are good reasons
to do so. First, IPTV providers want to guarantee that no control traffic clogs their
networks. Second, they want to be assured that propagation delays for join requests —
when users switch to a new TV channel — are modest, in order to minimise channel
change delay. So TV channels have to be distributed to as close to the users as possible.
But it is anyway possible to increase the resource (and associated energy) efficiency of
the network without jeopardising service quality, as I explain in Chapter 6.
GCHCH Ieik servixes
Alongside conventional TV, current IPTV providers often support additional features,
some of which are not offered by traditional TV services. For example, many add
sophisticated Electronic Program Guides and Set Top Boxes with extra functionality,
such as recording capabilities. Depending on the provider, IPTV users can also enjoy
many advanced features such as on-line gaming, chatting, and other web services on
their TVs.
In terms of the TV content itself, most systems today distribute Standard Definition
(SD) TV channels using MPEG-2, requiring 4 Mbps guaranteed bit rate per channel [5].
As optical fibre comes closer to the customer premises, higher capacities are becoming
available in the access link and several TV broadcasters are now offering High Definition
(HD), requiring around 20 Mbps per channel [29]. In the future, ultra high definition
systems are also expected. Panasonic, for instance, presented recently at the Consumer
Eletronics Show (CES) a 150-inch plasma TV set with 4k resolution [210] (Figure 2.6).
In the next decade, as prices plunge, this type of device may become common in our
living rooms, creating new market opportunities. These very high resolutions require
GCI Xhvllenges for Ieik proviyers GH
hundreds of Mbps per TV channel [83], significantly increasing the bandwidth demands
on the network and increasing its operational complexity. This may make current
static multicast systems prohibitive and justify the use of resource and energy-efficient
distribution schemes, as the ones proposed in Chapters 6 and 7 of this dissertation.
Figure 2.6: Panasonic’s 150” plasma TV presented at CES 2008 [210]
GCI Chvllenges for Ieik proviyers
In a fiercely competitive market as that of the telecommunications sector, IPTV service
providers have several challenges to address. These include financing difficulties, par-
ticularly in a time of economic crisis, the choice of the best business plan to supplant
competitors, and how to keep up with the recent technological advances and hurdles.
In this dissertation, I address the latter.
IP multicast offers the point-to-multipoint delivery mechanism necessary for the
efficient distribution of TV services. However, the original service model has some
issues that for some time have stalled the widespread use of multicast. In a paper
published a decade ago, Diot zt vlC [67] identified some of these issues:
GI Wvxkgrouny
• The multicast service model does not consider group management. This includes
authorisation for group creation and for transmission, billing policy and address
discovery.
• Security is also a problem. Authentication is not mandatory, and scalable key
management for encryption and data integrity is still an issue.
• Distributed multicast address allocation is another concern. Because the current
multicast address space is unregulated, nothing prevents applications to sending
data to any multicast address.
• Finally, there is no robust support for network management.
These problems still exist today but, as explained in the previous section, current
IPTV networks are provider-managed services. Being closed networks under the control
of a single entity eliminates the four problems identified by Diot zt vl. Still, other issues
persist. For example, IPTV offerings should match the level of quality of service that
customers are accustomed to from other TV service providers. Customers would not
tolerate poor quality of picture and sound. But the delivery of video is challenging. In
order to be successfully decoded in the Set Top Box, the video stream has to arrive at a
known and constant bit rate, in sequence, with minimal jitter or delay. The problem is
that IP networks are “best effort”, susceptible to lost or dropped packets as bandwidth
becomes scarce and jitter increases. In addition, video streaming requires high data
rates, so efficiency in distribution is a major concern. Also, the access network has
been a bottleneck until recently. These problems have been mostly solved:
1. The use of traffic prioritisation techniques and bandwidth reservation for IPTV
traffic assures that quality of service requirements are guaranteed in these closed
networks.
2. The emergence of scalable multicast protocols — in particular centre-based ones,
notably PIM-SM — provide the means for an efficient distribution of TV services.
3. The last mile bandwidth bottleneck has been broken in most developed countries,
with enhancements to DSL technology, and with the recent trend to bring fibre
closer to the home [201]. This, coupled with the progress of video codecs, such
as MPEG-2 and MPEG-4, to compress video content, and with the increased
processing power and storage capacity of (cheap) STBs, enables the distribution
of TV on current IP networks.
GCI Xhvllenges for Ieik proviyers G5
The fact that multicast IPTV has been fully deployed with success by several tele-
com companies — examples include AT&T, Telefonica, France Telecom, China Tele-
com, etc. — is evidence that these technological advances made it cost effective to
deploy and manage a multicast network.
But IPTV service providers still face important technological challenges. A relevant
one is still related to the provision of a level of service at least as good as its competitors.
This certainly includes offering new, added-value services, but it is also fundamental to
guarantee the quality of experience conventional TV users expect. IPTV service’s high
channel change delay, covered in Section 2.1.1, is still a thorn in IPTV provider’s side
in this respect. This is the first problem I explore in this dissertation, in Chapter 5.
Another technological challenge is to maintain an operationally cost and energy efficient
network in face of the evolution of IPTV services. This is the second problem I address.
As explained in Section 2.3, static multicast is inefficient. A dynamic multicast solution
also brings issues, such as network scalability and service quality, with more signalling
messages on the network, frequent router state changes requiring additional processing,
and an increase in channel switching delay (in certain periods some TV channels may
not be distributed close to the users requesting them). Chapters 6 and 7 are devoted
to this compromise between network efficiency and service guarantees.

Chvpter H
htvte of the vrt
The distribution of TV over IP networks provides an attractive business opportunity for
telecommunications service providers. The emergence of major players in this market is
an evidence of this fact. Early IPTV deployments have demonstrated that significant
technical challenges must be overcome to ensure the service is compelling to users
and competitive with other provider’s offerings. As a consequence, IPTV has posed
interesting research questions that were the subject of several papers over the past few
years. In this chapter I present research on IPTV and also on topics closely related to
the IPTV challenges identified in the previous chapter.
The chapter opens with research on IPTV network measurements. With the recent
deployment of IPTV networks a number of papers measuring and characterising IPTV
traffic have been published. The analysis of real IPTV workloads led to a clearer
understanding of how people watch TV and how this impacts the network. The findings
from these studies offered clues that led to some of the techniques I investigate in
Chapters 5, 6, and 7. The next three sections focus on the particular set of problems
addressed in this dissertation, namely, IPTV channel change delay and resource and
energy-efficiency of IPTV networks. First, I present a short survey on the research
done to date to mitigate the high channel change delays that occur in IPTV systems.
Afterwards, I devote a section to energy efficiency on networks, a topic that usually
goes under the label “green networking”. This chapter closes with a summary of work
on the integration of optical switching with electronic routing in IP networks. This is a
technique I explore in Chapter 7 with the objective of increasing the energy efficiency
of IPTV networks.
GM htvte of the vrt
HCF bevsurement of Ieik systems
The traditional methods used to assess TV viewing habits, employed by companies
such as Nielson Media Research [147]1, are based on monitoring of a sample of repre-
sentative users in order to extrapolate their behaviours to the entire population. The
bidirectional communication in IPTV systems offers new possibilities in this space, giv-
ing more visibility to viewers activities across an entire network. The availability of
IPTV workloads from large-scale IPTV systems can therefore be quite useful in under-
standing TV viewing habits. In order to comprehend how people watch TV and how
the network copes with the addition of this new service, a number of empirical studies
analysing IPTV traffic have been performed.
Cha zt vlC [42] presented the first analysis of IPTV workloads based on network
traces from one of the world’s largest IPTV systems2. The authors characterised the
properties of viewing sessions, channel popularity dynamics, geographical locality, and
channel switching behaviours. They discussed the implications of their findings and
explicitly mentioned the support needed for fast channel changes, a problem I address
in Chapter 5:
qihz yz“ign of v “y“tzm thvt “upport“ fv“t xhvnnzl “fiitxhing (CCC) i“ impzrvtivz to
woth improving u“zr zflpzriznxz vny minimi“ing thz impvxt in thz nztfiorkC7 [42]
By means of simulations using the same dataset, in [43] these authors consider
the limitations of current IPTV architectures based on static multicast distribution.
They proposed the integration of P2P distributed systems into the Set Top Boxes,
thus forming a cooperative P2P and IP multicast architecture. In this dissertation I
also look at the problem of using static multicast, by proposing a scheme that is a
compromise between static and dynamic multicast in Chapter 6.
Qiu zt vlC have also analysed channel popularity in the context of IPTV [161]. In
this paper the authors captured the channel popularity distribution and its temporal
dynamics. Later, the same researchers extended this work with the characterisation
and modelling of aggregate user activities in an IPTV network [160]. For both studies
they also used real data from an operational nation-wide IPTV system3. Their findings
overlap with those of [42]. In addition, in [160] the authors generalised the analysis
and developed a series of models for capturing the probability distributions and time-
1cizl“zn bzyiv gz“zvrxh pFILr mzv“urz“ mzyiv vuyiznxz“A proviying v fiiyz “zt of “tvti“tix“ on ik
vizfiing vny progrvm rvting“ wv“zy on v “vmplz of populvtionC
2ihzy u“zy v yvtv“zt from izlzfonixv’“ Ieik “zrvixz ImvgznioA thz “vmz yvtv“zt I u“z in thz xurrznt
“tuyyC
3In thi“ xv“zA Vi&i’“C
HCG V survey on texhniques to reyuxe Ieik xhvnnel xhvnge yelvy GN
dynamics of user activities. Lastly, they also combined these models to design an IPTV
workload generation tool.
Another empirical study of an IPTV network was presented by Mahimkar zt vlC [137].
The authors focused on characterising and troubleshooting performance issues on the
largest IPTV network in North America1. These researchers developed a diagnosis tool
capable of detecting and localising regions in the IPTV network experiencing serious
performance problems.
HCG V survey on texhniques to reyuxe Ieik xhvnnel xhvnge
yelvy
Channel change delay is one of the most severe problems affecting IPTV deployment,
and for that reason a good amount of research was done on this field so far. In this
section I present a brief survey of the proposed solutions to this problem in the litera-
ture.
HCGCF himple optimisvtions using pure multixvst
An obvious way to reduce zapping delay is to encode the video stream with a higher
frequency of I-frames. However, as explained in Section 2.1, such scheme would signif-
icantly increase the storage needs at the video server as well as the bandwidth needed
to offer the service. This is therefore not a practical solution.
More feasible solutions include the optimisation of channel streaming and playout.
Kopilovic and Wagner [129] have shown that it is possible to optimise channel streaming
with respect to initial buffering without increasing the bandwidth. This way they set a
limit on what is achievable by pure multicast without additional infrastructure. Kalman
zt vlC [118, 119] have shown how adaptive media playout — the variation of playout
speed of media frames depending on channel conditions — allows the client to buffer
less data, thus introducing less delay, for a given level of protection against buffer
underflow. In this scheme, the client varies the rate at which it plays out audio and
video according to the state of its playout buffer. When the buffer occupancy is below
a desired level, the client plays the media slowly, generating unnoticeable latency. This
latency is then eliminated with periods of faster-than-normal playout. This scheme is
similar to the adaptive piggyback techniques used to reduce I/O bandwidth in a Video
on Demand (VoD) server [4]. In fact, several techniques used in the past to reduce
1VgvinA Vi&iC
HE htvte of the vrt
VoD start-up delay [102] are now being transposed to mitigate channel change delay in
IPTV.
HCGCG eroxy server with woost strevms
Most commercial solutions to the channel change delay problem attempt to ensure that
an STB that is trying to join a new channel gets an auxiliary stream that starts with an
I-frame and then offers some kind of mechanism to switch over to the main multicast
stream. This is probably the most common fast channel change mechanism, and is
used, for example, by the Windows Media Platform [141]. This solution, illustrated
in Figure 3.1, requires the introduction of dedicated zapping servers in the network.
Simultaneously with the request to joint the multicast group, the Set Top Box (STB)
requests the channel from the zapping server (step 1 in the figure). The zapping server
then transmits a unicast burst with a delayed stream that starts with an I-frame (step
2). This stream is sent at a higher than usual bit rate, for the play-out buffer to fill
quickly. Thus the two main components of zapping delay are either removed — there
is no waiting for the I-frame, since this is the first to be received — or significantly
reduced — buffering time is lower. When an I-frame from the multicast flow finally
arrives, the STB terminates the unicast flow and switches to the multicast one (step 3).
In the figure the servers are co-located with the regional-metro routers, but they could
be placed at other locations. There is a trade-off between the number of servers needed
and the performance of the system. Fewer servers in the network means more requests
to respond per server, with a possible increase in response time (i.e., in channel change
delay).
This solution is expensive because it requires many dedicated servers to be added to
the network. To mitigate this problem Begen zt vlC [27, 28] proposed a unified approach
that can be used both to repair lost packets in real time and reduce the zapping delay.
So, if you already have a dedicated server to deal with lost packets, you may well use
the same for reducing channel change delay. The authors proposed to use the unicast
retransmission support of RTP [171] and RTCP [151], conventionally utilised to recover
lost packets, to accelerate channel changes. This feedback mechanism is used to provide
the key information needed by receivers to start processing the data prior to joining
the multicast session.
In the steady state, multicast reduces IPTV traffic volume. But channel surfing
disrupts this steady state when boost stream solutions are used in the network. The
network sends the unicast boost stream to make channel changes fast superimposing
HCG V survey on texhniques to reyuxe Ieik xhvnnel xhvnge yelvy HF
core IP network
IPTV
head-end
core
network
regional
network
metro
network
TV
channels
STB
access
network
Zapping
server
1
2
3
1
Figure 3.1: Zapping servers introduced in the IPTV network to mitigate the high
channel change delay
an additional demand on top of the steady state demand. This demand is propor-
tional to the number of users concurrently initiating a channel change event. Flash
crowds of channel changes (when channel changes are correlated, for example at the
completion of a popular program) place significant demands on the network and video
server resources. In [186] D. Smith analysed this problem by constructing a mathemat-
ical model to determine the bandwidth demand of a channel change mechanism where
unicast streams at higher than usual bandwidth are sent when viewers are changing
channels. This model quantifies the extra bandwidth consumed by channel surfing.
Smith looks particularly at commercial breaks since these periods are more disruptive
to the steady state demand. By assuming a simple exponential distribution for the time
HG htvte of the vrt
between channel changes, the author finds that the peak demand during a commercial
break is twice the steady state multicast demand.
Since these unicast solutions have this scalability problem, a few multicast-based
solutions were proposed recently. Sasaki zt vlC [176], for instance, propose the STB to
receive an additional multicast stream together with the original stream. This is simply
a delayed version of the original stream, resulting in the buffering time being halved.
If other viewers switch at the same time to the same channel, multicast suppresses
any duplication of packets. A similar proposal was presented by Banodkar zt vlC In
their proposal [23] the user joins a secondary multicast stream in association with the
multicast of the regular quality stream. This secondary stream is of lower quality
(it contains only I-frames, therefore it is not full motion video). During a channel
change event, the STB does a multicast join to this secondary stream, allowing the
user to experience smaller display latency. In the background, the playout-buffer of the
original full quality stream is filled, and when the play-out point is reached this full
quality stream is displayed and the transition is complete.
The most interesting multicast assisted zap acceleration system is probably the
one proposed by Bejerano and Koppol [30]. The main objective of their system is to
reduce the FID (First I-Frame delay): time until the following I-frame is received. For
this aim they also deploy an additional server in the provider network (they call it a
zapping accelerator, ZA). The ZA generates several time-shifted replicas of each TV
channel media stream, and each of these replicas is identified by a unique multicast
group. Also, all STBs are subscribed to a meta-channel, a low bandwidth multicast
group with information on the earliest replica with an I-frame for each TV channel, to
help the STB choosing the best one. When the user switches to a new channel, it joins
the multicast group of this replica. This way the zapping delay experienced is lower
and deterministic — it is bounded by the time-shift between two successive replicas
(more replicas mean a reduced time-shift, thus a lower zapping delay). To reduce
bandwidth consumption these replicas are sent only when there are users watching the
channel, and several users can be served simultaneously by the same replica (since they
are multicast streams). After a while the STB switches from the replica to the main
stream. As usual, to allow the STB to perform this migration transparently the replicas
are sent at a higher data rate. The fact that we need five or six of these higher bit rate
replicas to guarantee a low zapping delay will increase the bandwidth usage by each
TV channel in the network by about an order of magnitude. There is therefore a clear
tradeoff between the number of ZAs and bandwidth consumption.
These multicast solutions are more scalable since the server and network load de-
HCG V survey on texhniques to reyuxe Ieik xhvnnel xhvnge yelvy HH
pend not on the number of viewers zapping, but rather on the number of TV channels
the users are zapping to. This scalability enhancement allows a single server to serve
more users, which means fewer servers are needed in the network, reducing costs. Also,
with a multicast solution bandwidth consumption and server load are lower even during
flash crowds of channel changes.
HCGCH kiyeo xoying texhniques
Other type of solutions to the channel change delay problem include altering or ex-
tending the video coding techniques that were the subject of section 2.1. One such
technique is to include a picture-in-picture channel in the stream [65]. This channel
has a lower bit rate (and resolution) than the regular channel. It is constructed with a
small GOP size. When the STB tunes to the new channel it first tunes to this channel,
which is temporarily displayed until the STB has received an I-frame from the regular
channel and the play-out buffer has been filled to an acceptable level. A lower spacial
resolution is displayed during the zapping period, and the system requires a higher
bit rate because both channels are sent simultaneously. A similar scheme is presented
in [35], where Boyce and Tourapis proposed embedding a lower resolution stream into
a normal resolution one for each channel. The I-frames in the lower resolution scheme
occur more frequently, and when the user switches to this channel these streams are
decoded first, and only after a while is the normal stream used for playout.
One of the problems of the previous approaches is bandwidth inefficiency, so the
authors of [115] targeted this particular point in their proposal SFCS (Synchronisation
Frames for Channel Switching). In the schemes presented before, in order to switch
from the lower quality stream to a higher quality stream (bitstream switching) it is
necessary to wait for an I-frame, as these are the stream synchronisation points for
the client. As explained in section 2.1, the drawback of using I-frames is that, since
temporal redundancy is not exploited, they require a much larger number of bits than
P-frames at the same quality. For this reason, I-frames are an inefficient solution when
the actual requirements for stream synchronisation are taken into account. To be more
bandwidth-efficient the authors proposed using switching frames (SP and SI-frames)
instead of I-frames. The concept of switching frames was introduced in [121]. The main
feature of SP-frames is that identical frames can be reconstructed even when different
reference frames are used for their prediction. This allows them to replace I-frames
in applications such as bitstream switching, as in this case. Albeit providing similar
functionality, SP-frames have significantly better coding efficiency than I-frames: since
HI htvte of the vrt
they utilise motion-compensated predictive coding, they require fewer bits than I-frames
to achieve similar quality. The same authors also proposed an extension to SFCS [114]
to be compatible with encoding/decoding systems that do not support SI/SP-frames.
In [117] Joo zt vlC proposed an algorithm to control channel zapping time by adjust-
ing the number of broadcast IPTV channels that are distributed close to users and the
number of I-frames inserted into each channel, based on the user’s channel preference
information. They thus consider two variables — broadcasting channel distribution (po-
sitioning channels according to their popularity) and video encoding structure (adding
extra I-frames to the normal video frames to decrease video decoding delay). With
their algorithm they achieve an effective trade-off between channel zapping time and
network utilisation.
The main problem of all these schemes is the increase in encoder complexity. Besides
this, most require additional bandwidth. In addition, the user will experience a brief
lower quality period immediately after zapping, and the transition from a low to a
high-resolution channel can frequently cause undesirable glitches.
HCGCI ereyixtive preBjoining of ik xhvnnels
As explained in Chapter 2, in IPTV systems typically only the channel the user is
watching is delivered to the Set Top Box at any one time, due to bandwidth limitations
in the access network. When a user switches to a new channel, the STB has to issue a
new channel request towards the network. The fact that the channel the user switched
to is not available in the STB is the main reason for a high channel change delay:
the STB has to join the multicast group from this channel, synchronise with the video
content and buffer some packets before play-out. In predictive pre-joining schemes, each
STB simultaneously joins additional multicast groups along with the one requested by
the user, thus anticipating future user behaviour. These schemes are thus based on the
prediction of the next TV channels the user will switch to. If the prediction is right,
the user will experience a small zapping delay, as the channel switched to is already
synchronised in the STB.
The first paper proposing pre-joining of TV channels was by Cho zt vlC [51]. In
their proposal the additional channels are simply the channels adjacent to the channel
being watched. The main problem of this work was that no evaluation was given. The
paper offered a mere description of the idea. Furthermore, without modifications their
scheme would be very inefficient. The adjacent channels would be sent continuously
with the requested channel. So, in the periods when the user is settled in a channel (i.e.,
HCH Green networking H5
not zapping), the adjacent channels would be transmitted in the access link consuming
precious bandwidth.
Two other papers [81, 189] proposed a similar scheme, but solved the two problems
of the original. First, they considered delivering the adjacent channels for a finite period
only, thus their schemes are bandwidth-efficient. Also, they zvvluvtzy their proposals by
developing an analytical model to investigate the performance of each of their schemes.
The main problem of these two papers is that some of the assumptions they made
in building their simple models have been proved wrong recently: by analysing real
IPTV datasets, recent studies [160, 166] have shown that channel surfing behaviour
should not be modelled with the simple Poisson processes the authors of [81, 189] (and
others) used. The constant-rate Poisson models generally used as workload model are
not capable of capturing the high burst of channel switches at particular periods1. For
evaluation of IPTV studies it is therefore important to use actual IPTV trace data or
reliable models based on empirical data, such as the one proposed by Qiu zt vlC [160].
Recently more sophisticated pre-joining schemes have been considered. Oh zt vlC [150]
presented an hybrid scheme combining pre-joining and reordering. The authors con-
sidered two pre-joining schemes: a first where the adjacent channels, as above, are
pre-joined, and a second where the most popular channels are pre-joined. They then
combined these with a channel reordering scheme the same authors had proposed be-
fore [131], where popular channels are clustered together in the linear search sequences.
Another recent paper proposing a pre-joining method for the same purposes was pre-
sented by Lee zt vlC [130]. Their scheme is based on both button and channel preference.
The authors also included a method to determine the most efficient number of channels
to pre-join. A small number of channels is pre-joined during viewing periods, with
more channels being pre-joined during zapping periods. In these two papers simple
mathematical models were also used for evaluation, with the same problems described
before.
I target this category of techniques — predictive pre-joining of TV channels — in
Chapter 5 of this dissertation.
HCH Green networking
Since the seminal paper by Gupta and Singh [93], presented at SIGCOMM in 2003,
the subject of grzzn nztfiorking has received considerable attention. In recent years,
valuable efforts have been devoted to reducing unnecessary energy expenditure. Big
1I rzturn to thi“ i““uz in thz nzxt xhvptzr to provz thi“ fvxtC
H6 htvte of the vrt
companies such as Google, Microsoft, and Amazon, are turning to a host of new tech-
nologies to reduce operating costs and consume less energy [122]. Google, for example,
is planning to operate its data centres with a zero carbon footprint by using, among
other things, hydropower, water-based chillers, and external cold air to do some of the
cooling.
Several approaches have been considered to reduce energy consumption in networks.
These include:
• The design of low power components that are still able to offer acceptable lev-
els of performance. For example, at the circuit level techniques such as Dynamic
Voltage Scaling (DVS) and Dynamic Frequency Scaling (DFS) can be used. With
DVS the supply voltage is reduced when not needed, which results in slower op-
eration of the circuitry. DFS reduces the number of processor instructions in a
given amount of time, thus reducing performance. These techniques can reduce
energy consumption significantly. Zhai zt vlC [219] have shown that theoreti-
cally the power consumption decreases cubically when DVS and DFS are applied
jointly. As with a reduced frequency the time to complete a task increases, the
authors show this is translated into an overall quadratic reduction in the energy
to complete a task.
• Consuming energy from renewable energy sources sites rather than incurring in
electricity transmission overheads [68], thus reducing Cd2 emissions.
• Designing new network architectures, for example by moving network equipment
and network functions to strategic places. Examples include placing optical am-
plifiers at the most convenient locations [195] and performing complex switching
and routing functions near renewable sources [68].
• Using innovative cooling techniques. Researchers in Finland, for instance, are run-
ning servers outside in Finnish winter, with air temperatures below -20 ◦C [155].
• Performing resource consolidation, capitalising on available energy. This can be
done via traffic engineering, for instance. By aggregating traffic flows over a subset
of the network devices and links allows others to be switched off temporarily or be
placed in sleep mode [93]. Another way is by migrating computation, typically
using virtualisation to move workloads transparently [33, 61]. Computation is
migrated from several lightly loaded devices or servers to a consolidation server,
and then the equipment that is freed up can be turned off.
HCH Green networking HL
• Reducing router processing, for example by switching transit traffic1 at the optical
layer [14]. This technique is the basis of the proposal I present in Chapter 7.
In the past few years numerous researchers have used these techniques to build
greener networks. In the rest of this section I present a brief overview of the state of
the art on this topic.
HCHCF Funyvmentvls
Before attempting to reduce energy consumption, it is important to know the funda-
mentals in order to identify where significant savings can be obtained. In a series of
two papers, Ronald Tucker explored the fundamental limits on energy consumption in
optical communication systems and networks. In Part I [194] the author focused on the
lower bound on energy in transport systems. Among other results, he concluded that
it is possible to minimise the total energy consumption of an optically amplified system
by locating repeaters strategically. In Part II [195] Tucker explored the lower bound
on energy consumption in optical switches and networks, confirming a previous find-
ing [14] that the energy consumption of the switching infrastructure is larger than the
energy consumption of the transport infrastructure. Still on the fundamentals, Baliga
zt vlC [15] suggested that the ultimate capacity of the Internet might eventually be con-
strained by energy density limitations and associated heat dissipation considerations
rather than the bandwidth of the physical components.
Energy-awareness has increased in proportion with the emergence of recent studies
that quantified the energy consumption of networks. In [14], for example, the authors
presented a network-based model of power consumption in optical IP networks and used
this model to estimate the energy consumption of the Internet. They estimated that
the Internet currently consumes around 0.4% of electricity consumption in broadband-
enabled countries, but that this figure is on the rise. Other studies have suggested an
important increase of core network energy consumption. For instance, Tucker zt vlC [17,
19, 197] developed simple energy-consumption models in a series of papers and reached
the overall conclusion that at low bit rates power consumption is dominated by the
access network. However, as access rates to users increase, the energy consumption in
routers, particularly core routers, will become significant and eventually dominate.
1irvffix not yz“tinzy to thz noyz unyzr xon“iyzrvtionC
HM htvte of the vrt
HCHCG co workT ihenA sleep
One of the most common techniques to save energy is to shutdown network equipment
(or some of its constituent components) whenever possible. In the pioneering work
on green networking [93] the authors discussed the impact on network protocols of
saving energy by putting network interfaces and other router and switch components
to sleep. They considered changing routes during low activity periods so as to aggregate
traffic along a few routes only, while allowing devices on the idle routes to sleep. In this
position paper the authors concluded that sleeping was indeed a feasible strategy. Later,
the same authors have also examined the feasibility of putting various components on
LAN switches to sleep during periods of low traffic activity [92]. Based on traffic
collected in their LAN, they concluded that sleeping is feasible in a LAN environment
with little impact in other protocols, thus enabling energy savings.
Gupta and Singh continued their work on energy conservation in Ethernet LANs,
and in [94] they proposed methods that allow for detection of periods of inactivity
in these networks to obtain energy savings with little impact on loss or delay. Using
real-world traffic workloads and topologies (from Intel Enterprise network), Nedevschi
zt vlC [145] have also shown that such simple schemes for sleeping can offer substantial
energy savings.
Other research works follow the same line. Chiaraviglio zt vlC [50] considered a real-
istic IP network topology and evaluated the amount of energy that can be potentially
saved when nodes and links in the network are turned off during off-peak periods. They
also proposed a simple algorithm to select the network equipment that must be pow-
ered on in order to guarantee the service. Fisher zt vlC [77] developed and evaluated
techniques to save energy in core networks by selectively powering down individual
cables of large bundled links during periods of low utilisation. Idzikowski zt vlC [108]
estimated potential energy-savings in IP-over-WDM networks achieved by switching
off router linecards in low-demand hours. All these works show that it is possible to
achieve significant energy savings using such simple techniques.
HCHCH aittle workT ihenA slow yownO Vyvptive aink gvte
It is recognised by the research community devoted to green networking that for systems
and networks to be energy-efficient, energy proportionality should become a primary
design goal [24]. An efficient device should consume energy proportionally to its output
or utility. Unfortunately, most equipment is not energy-proportional. Fortunately, a
serious effort is being made in the community to change the current situation.
HCH Green networking HN
A common example is the Ethernet. Ethernet power consumption is independent of
link utilisation. Idle and fully utilised Ethernet links consume about the same amount
of power. As the average utilisation of desktop Ethernet links is very low, in the range
of 1 to 5 percent [52], and as this situation is likely to persist [149], the current state of
affairs is undesirable. A way to improve Ethernet so that energy use is proportional to
link utilisation is Adaptive Link Rate (ALR). ALR, proposed by Gunaratne zt vlC [90],
is a means of automatically switching the data rate of an Ethernet link to match link
utilisation. Based on simulation experiments using actual and synthetic traffic traces
these authors have shown in [91] that an Ethernet link with ALR can operate at a
lower data rate for over 80% of the time, yielding significant energy savings without
compromising the quality of service.
HCHCI lvnt to sleepT ihenA yelegvteO eroxying
Desktop computers in enterprise environments consume a lot of energy in aggregate
while still remaining idle much of the time. The question many researchers have asked
in the past few years is how to save energy by letting these machines sleep while avoiding
user disruption. To reduce energy waste by idle desktops the typical approach is to
put a computer to sleep during long idle periods, with a proxy employed to reduce user
disruption by maintaining the computer network’s presence at some minimum level.
The proxy can be co-located within the host (e.g., on an Ethernet NIC), or in another
device (e.g., a LAN switch). The problem with this approach is the inherent trade-
off between the functionality of maintaining network presence and the complexity of
application specific customisation.
The simpler such mechanism is the Wake on LAN technology [10]. This is a mid-
1990s industry standard that makes it possible for an Ethernet adapter to wake-up
a sleeping desktop computer using a specially defined packet (a “magic packet”). A
more sophisticated approach was proposed by Christensen zt vlC [52]. The authors
propose a proxying Ethernet adapter that can wake-up a desktop computer in sleep
mode when its resources are needed and otherwise handle routing protocol messages
without waking up the computer. Thus the computer maintains its presence on the
network without being fully powered on at all times. An in-depth evaluation of the
potential energy savings and the effectiveness of proxy solutions was performed by
Nedevschi zt vlC [144]. They considered two types of proxy. A simple one that performs
automatic wake up triggered by a filtered subset of the incoming traffic, and a more
elaborate one which incorporates application-specific stubs that allow it to engage in
IE htvte of the vrt
network communications on behalf of applications running in the machine that is now
sleeping. Other examples of application-specific stubs exist for BitTorrent [11] and
Gnutella [116]. A proxy prototype was also presented recently by Agarwal zt vlC [3]:
Somniloquy. This proxy is an augmented network interface that allows PCs in sleep
mode to be responsive to network traffic. Somniloquy achieves this functionality by
embedding a low power secondary processor in the network interface.
Another approach to save desktop energy is by virtualising the user’s desktop com-
puting environment as a virtual machine, and then migrating it between the user’s
physical desktop machine and a VM (Virtual Machine) server. This is the idea behind
Litegreen [61]. With this system, the user’s desktop environment is “always on”, main-
taining network presence even when the user’s machine is switched off. The idle desk-
tops are consolidated on the server. A similar scheme is proposed by Bila zt vlC [33].
While Litegreen migrates the VM’s entire memory state to the consolidation server,
these authors propose partial migration, with only the working set of the idle VM
being migrated. This allows an almost instantaneous and low energy cost migration.
HCHC5 Chvnge pvrvyigmO energyBvwvre infrvstruxtures
A consensus exists in the green networking community: it is necessary to introduce
energy-awareness in network design. Certainly, this has to be achieved without com-
promising either the quality of service or the reliability of the network. A good amount
of effort has been therefore devoted to the design of novel energy-aware infrastructures.
In some cases, new networking paradigms are being explored.
Until recently power consumption has been dominated by access networks [17]. As
a consequence, some research has been devoted to this specific part of the network.
To achieve power-savings in DSL networks, for example, Cioffi zt vlC [55] and Tsi-
aflakis zt vlC [192] proposed dynamic spectrum management (DSM). DSM improves
DSL networks by tackling the crosstalk problem1. Typically, DSM algorithms focus
in maximising data rates. Tsiaflakis zt vlC [192] extended these algorithms by incor-
porating energy efficiency as an objective. Panarello zt vlC [153] followed a different
approach. They proposed a combined congestion control and rate-adaptation scheme
for Internet access nodes which allows them to reduce energy consumption.
The energy consumption of big data centres (“clouds”) has also been a hot topic
of research. Heller zt vlC [98] presented a power manager and optimiser, ElasticTree,
which dynamically adjusts the set of active network elements — links and switches —
1ihz xro““tvlk i“ vn zlzxtromvgnztix intzrfzrznxz gznzrvtzy wy yiffzrznt linz“ opzrvting in thz “vmz
xvwlz wunylzC
HCH Green networking IF
to satisfy changing data centre traffic loads. This introduces energy proportionality
into the network even though individual network devices are not energy-proportional.
Most studies on green cloud computing focus on the energy consumed in the data cen-
tre. However, the increase in network traffic that results from moving computation to
the cloud may also have an impact in energy consumption. This is the topic of [18].
In this paper, Baliga zt vlC presented an analysis of energy consumption in cloud com-
puting. By building mathematical models based on power consumption measurements
and published specification of representative equipment, the authors showed that en-
ergy consumption in transport and switching can represent a significant percentage of
the total energy consumption, and thus should be carefully considered. The authors
argued that under some circumstances cloud computing can consume more energy than
conventional computing.
A change in network architecture paradigms from host-oriented to content-centric
networking (CCN) created new possibilities for energy-efficient content dissemination.
The main difference between a CCN router and an IP router is that the former sup-
ports name-based routing and caching for content retrieval [112]. These content-caching
capabilities can significantly reduce redundant content transmission and consequently
avoid energy waste. A study by Lee zt vlC [132] shows that CCN is capable of out-
performing conventional Content Distribution Networks (CDNs) and P2P networks in
terms of energy-efficiency. The authors used a publicly available traceroute dataset to
validate their claim.
Several proposals to change the current client-server paradigm of Video on Demand
(VoD) services have been proposed recently. Valancious zt vlC [200], for example, pro-
posed a new way to deliver VoD based on the Nano Data Center (NaDa) platform.
NaDa uses ISP-controlled home gateways (Set Top Boxes) to provide computing and
storage and adopts a managed peer-to-peer model to form a data centre infrastructure
with these devices. The authors claim to achieve energy savings by analysing a set of
empirical VoD access data. The fact that they are reusing already committed baseline
power on underutilised gateways and that they avoid cooling costs is the justification
for these results. Feldmann zt vlC [72] also explored the energy trade-offs between P2P,
data centre architectures and CDNs in the context of Internet TV, and reached a some-
what different conclusion. Their results showed that P2P, albeit capable of reducing
the power consumption of the service provider, increases the overall energy consump-
tion. The reason is that P2P applications push the energy use out of the data centres
and into the homes of content consumers (thus migrating the problem). Another paper
on energy efficiency of VoD networks is by Baliga zt vlC [16]. The authors built an
IG htvte of the vrt
energy consumption model for these networks based on specifications of commercial
equipment. Their main conclusion is that to have an energy-efficient architecture pop-
ular new-release movies should be widely replicated throughout the network (as power
consumption of transmission dominates over storage), and progressively withdrawn to
fewer data centres as their usage declines (as power consumption of storage becomes
dominant).
The last type of proposals in this category includes changes to current network
applications. As an example, Blackburn and Christensen [34] proposed changes to the
BitTorrent protocol to make it “greener”. The current protocol requires clients to be
fully powered on to be participating members in the P2P overlay network (a “swarm”
in BitTorrent). The authors proposed simple changes such as including long-lived
knowledge of sleeping peers and a new wake-up semantic. This allows clients to sleep
when not actively downloading or uploading, yet still be responsive swarm members.
HCHCK hhifting to svveO irvffix Engineering
Novel traffic engineering algorithms and techniques have recently been proposed with
the aim of reducing energy consumption. Shifting traffic around allows specific equip-
ment to be switched off and computation to be performed in greener locations.
Restrepo zt vlC [169] proposed a novel energy reduction approach that takes load-
dependent energy consumption information of communication equipment into account
when performing routing and traffic engineering decisions. A similar work was done
in [202], where the authors assumed the availability in the near future of networking
hardware in which an interface can operate at various sending rates. The main idea
of the technique they proposed is to distribute traffic across alternative paths in a way
that maximises energy savings. In [54] Cianfrani zt vlC proposed a novel network-
level strategy based on a modification of current link-state routing protocols, such as
OSPF. According to this strategy, IP routers are able to power off some network links
during low traffic periods. The authors proposed a modified Dijkstra Shortest Path
First algorithm that detects links to power off. Switching between the active and sleep
modes consumes considerable energy and time, which motivated Andrews zt vlC [13] to
consider the scheduling problem jointly with the routing problem. Routing determines
the path each connection should follow, while scheduling decides the active periods for
each network element. By combining the two problems the authors were capable of
simultaneously minimising energy and end-to-end delays.
From an environmental point of view, the objective of green networking is to min-
HCH Green networking IH
imise greenhouse gases emissions. Enforcing the use of renewable energy is an important
step in this direction. Dong zt vlC [68] proposed a novel approach with this aim. Similar
to the work by Shen and Tucker [180], the authors of [68] developed efficient approaches,
ranging from Mixed Integer Linear Programming (MILP) models to heuristics to min-
imise energy consumption of IP over WDM networks. But Dong zt vlC go a step further
and also attempt to reduce Cd2 emissions by maximising the use of renewable energy
sources in the network. The idea is to ship information to distributed renewable en-
ergy locations, processing and switching remotely instead of transporting the energy
generated by renewable sources.
HCHCL avxking informvtionT bevsure it
Device specification data-sheets of current network equipment do not include compre-
hensive energy consumption values. They report merely maximum rated power. This
value is insufficient to understand the actual energy consumption of the networking
device under different configurations or traffic loads. Unfortunately, due to a lack of
empirical studies, much of the research on green networking, including many of the pa-
pers referred in this section, is based on these figures. To try to alleviate this problem,
Chabarek zt vlC [44] measured the power demands of two widely used Cisco routers
and created a generic model for router power consumption. In addition, the authors
proposed optimisation techniques to determine the optimal system configurations that
minimise power consumption while preserving performance requirements. A related
work was presented in [136]. Mahadevan zt vlC presented a power measurement study
of a variety of networking gear, and also proposed a novel network energy proportional-
ity index. More recently, Sivaraman zt vlC [184] proposed a fine-grained profile of energy
consumption on the NetFPGA platform1. By using a high-precision hardware-based
traffic generator and analyser, and a high-fidelity digital oscilloscope, the authors de-
vised a series of experiments allowing them to quantify the per-packet processing energy
and per-byte energy consumption of a NetFPGA card.
To be able to evaluate the performance of energy-aware networks it is important to
have a common framework for measuring and reporting the energy consumption of a
network. With this goal in mind, Bianzino zt vlC [32] compared and contrasted various
energy-related metrics and defined a taxonomy of green networking metrics, which is
1ihz czt[eGV i“ v lofiBxo“t rzxonfigurvwlz hvryfivrz plvtform optimi“zy for highB“pzzy nztfiorkingC
It xon“i“t“ of v fully progrvmmvwlz milinx [eGV wv“zy xorz fiith four Gigvwit Ethzrnzt intzrfvxz“A vny
funxtion“ v“ vn Ie routzrC aikz xommzrxivl routzr“A thz zntirz yvtvpvth i“ implzmzntzy in hvryfivrzC
ihz czt[eGV xvn thu“ “upport full Gigvwit linz rvtz“ vny hv“ lofi proxz““ing lvtznxy pFIKrC
II htvte of the vrt
probably an important first step to reach a consensus in the research community that
is devoted to these matters.
For more detailed surveys on green networking I forward the reader to [31] and [220].
Bianzino zt vlC’s survey [31] is computer networking-oriented, whilst the other, by Zhang
zt vlC [220], is optical networking-oriented.
HCI From elextronixs to optixs
The integration of optics and electronics in IP networks has been a hot topic of research
in the past decade. The technique I propose in Chapter 7 to reduce the energy footprint
of IPTV networks is an example of such integration. In this section, I review some
research done on this subject. Additionally, in Appendix A I present some topics that,
despite their orthogonality to the proposal presented in that chapter, are nonetheless
closely related. Namely, in the appendix I address optical multicast, traffic grooming,
and aggregated multicast.
HCICF dptixs vs elextronixsO there is room for woth
Optical technologies inherently high bandwidths greatly exceed the bandwidth of any
conceivable electronic device. The information-carrying capacity of optics is thus well
beyond the capabilities of electronics. This imbalance between optics and electronics
is sometimes referred to as the “electronic bottleneck”. By realising this fact, in the
past decade several research groups have analysed the problem of integrating optical
technology inside routers to scale its capacity and reduce its power consumption. An
example is the work by Keslassy zt vlC [123]. In their paper the authors identified an
optical switch architecture with predictable throughput and scalable capacity — the
Load-Balanced switch proposed by Chang zt vlC in [46] — and extended it in order to
solve the problems that made the original switch unsuited for a high-capacity router.
To take full advantage of the capabilities optics can offer, the final goal is to build
an all-optical router. At present, however, there does not appear to be a compelling
case for replacing electronic routers with all-optical packet switches [193]. The key
challenge in finding a technically feasible solution to optical packet switching is the lack
of an adequate optical buffering technology. The most commonly used optical buffers
are based on fibre delay lines, which are physically very large and inflexible. Electronic
RAM is still the most attractive choice due to its small size and low power consumption.
Another problem is the still immature optical signal processing technology. Only very
simple signal processing, such as wavelength conversion or regeneration, is amenable
HCI From elextronixs to optixs I5
to photonic implementation [99, 100]. It seems therefore clear that electronics will
continue to be the technology of choice for high-performance signal processing and for
buffering in the future.
But can optical technology help in reducing the energy footprint of networks? In
terms of processing capabilities, integrated nonlinear optical circuits still consume sig-
nificantly more energy than CMOS in all but the very simplest of circuits [196]. This
is mainly because in CMOS most of the switching energy is consumed during bit tran-
sitions, while photonic devices rely on optical non-linearities that require an ongoing
supply of power. So it does not seem a good option in this respect. On the other hand,
in terms of routing and switching capabilities, techniques such as optical bypass can be
an interesting energy-friendly option [99]. This is the technique I propose in Chapter
7 to reduce the energy consumption of IPTV networks. I briefly explain its rationale
in the following.
In core routers, power consumption is dominated by forwarding and cooling. Ad-
dress resolution and packet forwarding consume approximately 40% of router power [15].
As most of the traffic handled by a router is transit traffic [175], such electronic pro-
cessing is wasteful. Optical bypass can eliminate this expensive high-speed electronic
processing at intermediate nodes, and thus save energy. Without bypass, all lightpaths1
incident to a node must be terminated, i.e., all the data carried by the lightpaths has
to be electronically processed and forwarded by IP routers. With optical bypass, traffic
not destined for a given node is placed onto a WDM wavelength that is not processed
by that router. This can be accomplished by placing a WDM circuit-switched opti-
cal cross-connect (OXC) between the router and the incoming optical port so as to
direct channels not destined to that router directly to the node output [175]. Figure
3.2 illustrates such node in a simplified manner. With optical bypass the traffic tran-
siting a node can therefore remain in the optical domain, as opposed to undergoing
costly Optical-Electrical-Optical (OEO) conversions and per-packet inspections. This
can significantly save the number of IP router ports and consequently reduce energy
consumption [180]. As an aside, OEO conversions are also undesirable as they off-
set the high-speed of the optical transport. There are, however, limitations on the
use of optical bypass. The most important is its coarse granularity. OXCs switch
at the wavelength-level. This inflexibility in switching granularity can cause waste of
bandwidth. Internet traffic has many small and diverse flows which emphasises the
importance of resource sharing. The lack of multiplexing gain of all-optical switching
1In vn vllBoptixvl nztfiorkA v lightpvth i“ vn optixvl pointBtoBpoint xonnzxtion from v “ourxz to v
yz“tinvtionC
I6 htvte of the vrt
is therefore a disadvantage that should be considered.
core router
low end
routers
OEO
converters
OXC
WDM links optical bypass
no optical bypass
Figure 3.2: Optical bypass-enabled network node
HCICG Hywriy vrxhitexturesO the west of woth worlys
Optical cross connects relieve electronics from processing and switching. The problem,
as explained before, is that due to its coarse granularity, bulk transport in optics can be
bandwidth inefficient, especially for bursty traffic. With electronic switching the packets
or flows can be processed at a much finer granularity. Smartly combining the strengths
of optics and electronics seems therefore to be a good option. This type of hywriy
architecture (also called multi-granular [218] or translucent [179], among other nomen-
clatures) that represents a compromise between all-electronic and all-optical switching
has been the subject of interesting research in the past few years.
S. Aleksic [7], for instance, examined different switching and routing architectures
based on both pure packet-switched and pure circuit-switched designs by assuming
either all-electronic and all-optical implementations. The author concluded that to
build energy efficient networks a kind of dynamic optical circuit switching should be
HCI From elextronixs to optixs IL
used within the core network together with an efficient flow aggregation at edge nodes.
Enablers for this type of networks are novel hybrid optical cross-connect architectures
combining slow (millisecond regime) and fast (nanosecond regime) switching elements,
as the one proposed recently by Zervas zt vlC [218]. This equipment is able to switch
at the fibre, waveband, wavelength and sub-wavelength granularities. Several other
examples of hybrid architectures exist in the literature [37, 82, 88, 101, 177].
An example of a routing scheme that makes use of these hybrid architectures is
the work by Huang and Copeland [105]. The authors proposed a hybrid wavelength
and sub-wavelength routing scheme that can preserve the benefits of optical bypass
for large traffic flows and still provide multiplexing gain for small traffic flows. The
idea is to route traffic demands with large granularity using wavelength routing and
those with small granularity using sub-wavelength routing. They therefore propose a
“dedicated” set of wavelength channels to be optically switched and a “shared” one
to be electronically routed. This scheme is similar to what I propose in Chapter 7 for
IPTV systems.
The integration of optical circuit-switching techniques with electronic packet-switching
requires a unified control plane. This is an essential component in the evolution of inter-
operable optical networks. Generalized Multiprotocol Label Switching (GMPLS) [140],
the emerging paradigm for the design of control planes for OXCs, is a promising tech-
nology by providing the necessary bridges between the IP and optical layers [152].
GMPLS extends the Multiprotocol Label Switching (MPLS) [172] control plane to
encompass several switching granularities, from packet and layer 2 switching to wave-
length and fibre switching. The development of GMPLS required modifications to
current signalling [22] and routing [127] protocols. Extensions to RSVP-TE [154] and
OSPF-TE [126] in support of GMPLS were already standardised.

Chvpter I
bethoyology vny yvtvset
The ideas I propose and analyse in this dissertation are evaluated by means of trace-
driven analysis. It is widely accepted [113] that a thorough evaluation using real work-
loads enables the assessment of future network architectures with an increased level of
confidence. This chapter opens with an explanation of the motivation for the chosen
methodology. Then, I describe the dataset used in this study, detailing how the data
were collected, cleaned and treated. The chapter ends with an analysis that aims to
validate the dataset. This validation consists in analysing specific characteristics of the
data trace and contrasting the results obtained with those from a similar study made
by other researchers using a different dataset.
ICF bethoyology
The research community working on IPTV systems has relied upon hypothetical user
models which are sometimes different from reality and can lead to incorrect estimation
of system performance. As I already mentioned in the previous chapter, constant-rate
Poisson models are generally used as workload model for these systems. Examples
include [189], [81], [150], [130], among others. Unfortunately, this model does not
capture IPTV user behaviour well. Users switch channels more frequently than this
simple model predicts. This fact was proved by Qiu zt vlC [160] recently. These re-
searchers have characterised and modelled user activities in an IPTV network. They
used real data from an operational nation-wide IPTV system1. Based on their analysis
the authors developed a series of models for capturing the probability distributions
and time-dynamics of user activities. They show that the simple mathematical models
1Vi&i’“C
5E bethoyology vny yvtvset
generally used in these studies are not capable of capturing the high burst of channel
switches at around hours boundaries, and are thus not good models.
By analysing the dataset used in this dissertation1, I also observe this fact. In Figure
4.1 I demonstrate, by means of an example, the problem of using a simple Poisson
distribution as a mathematical model to represent the behaviour of IPTV users. The
figure presents the Cumulative Distribution Function of the number of channel switches
during one-minute periods (a zvpping pzrioy, according to [42]). The analysis was done
on the whole dataset (containing all channel switching events from 255 thousand users
over a six month period). In the figure I compare the empirical data with a Poisson
distribution with parameter  equal to 1.948.
0.0
0.2
0.4
0.6
0.8
1.0
Poisson model
Empirical data
10 20 30 40 50 60
Number of channel switches during zapping
C
D
F
Figure 4.1: Number of channel switches in zapping mode
As can be seen, the Poisson model is conservative in terms of the number of channel
switches a user performs during zapping periods. For example, the probability of a user
making five channel switches or more in a one-minute period is negligible when using
the Poisson distribution. But in fact by observing the empirical data one can conclude
that there is a 20% probability of a user switching channels five times or more during
a zapping period. This observation has important consequences for the current study.
For instance, the fact that users enjoy zapping more than the Poisson model predicts
is a stronger argument for the use of the scheme evaluated in Chapter 5.
The lack of an acceptable mathematical model for IPTV user behaviour2 and the
1lhixh i“ yz“xriwzy in “omz yztvil in hzxtion ICGC
2ihz rzvli“tix moyzl propo“zy wy fiu cr al, pFKEr fiv“ not vvvilvwlz fihzn thz wulk of thi“ “tuyy fiv“
ICG Dvtvset 5F
availability of an IPTV trace are the two reasons why I opted for trace-driven analysis
as the methodology used to evaluate the schemes proposed in this dissertation. This
IPTV trace from Telefonica is used as input to the analysis performed and presented
in chapters 5, 6 and 7. In those chapters I explain the precise methodological details
of each particular experience.
ICG Dvtvset
I was fortunate to obtain a collection of IPTV channel switching logs from an IPTV
service — Imagenio — offered by an operational backbone provider, Telefonica. Ima-
genio is a commercial, nationwide service, offering 150 TV channels over Telefonica’s
IP network. The access links use ADSL technology and the network is composed of
680 DSLAMs distributed along 11 regions. To give an idea of the scale of the dataset,
the 700GB trace spans six months and records the IGMP messages on the channel
changes of around 255 thousand users. The number of daily channel switchings clocks
13 million on average.
ICGCF Dvtv xollextion
I should start by clarifying that I was not responsible for trace collection. This process
was executed by engineers at Telefonica. In this subsection I describe in some detail
the data collection process. The information included here is the result of several
discussions with researchers and engineers from Telefonica R&D1.
As I explained in Chapter 2, whenever an IPTV user switches to a new TV channel,
two IGMP messages are generated by the Set Top Box (STB) and sent towards the
network: an IGMP leave request from the current TV channel, and an IGMP join
request to the TV channel the user is switching to. In Telefonica, as in most IPTV
networks, all channels are distributed continuously to all DSLAMs. This maintains the
channels as close to the users as possible to help reduce channel change delay and avoids
signalling messages in the IP network. The leave and join messages therefore arrive at
the DSLAM, which then distributes the corresponding TV channel to the STB that
requested the change. To collect the traces the DSLAMs were instrumented to send vll
IGMP join and leave requests sent by vll STBs to a particular server in its region: a
wzing rzvli“zyC
1I fiouly likz to zxprz““ my grvtituyz in pvrtixulvr to evwlo goyriguzzA Jvvizr Wznito vny Enriquz
jrrzv from izlzfonixv g&DA vny to bzzyoung Xhv from KVIhi =intzrn vt izlzfonixv g&D vt thz timz
of yvtv xollzxtion) for kinyly “hvring vll thz yztvil“ vwout thz proxz““C
5G bethoyology vny yvtvset
loxvl vrzv “zrvzr.
Figure 4.2 pictorially illustrates the data collection process. Every time the DSLAM
received an IGMP leave or join message from an STB it would forward this message to
the local area server. One local area server served many DSLAMs, and some regions
had more than one such server (for example, Madrid and Barcelona had four local area
servers each). Each local area server then recorded every message received from the
DSLAMs of its area into a log. The server was also running a script that periodically
sent all log files to a central data collection server using a crontab1. There was only
one central server keeping all logs.
Concerning the reliability of data collection, three sources of errors should be taken
into account: errors in communication, problems/failures in the network elements, and
possible lack of synchronisation. Concerning the first problem, all log messages were
sent over UDP, so there was indeed the possibility of messages being lost. Second,
log collection was regarded as a low priority process in Imagenio2. Therefore, in the
event of a DSLAM processor overload, for instance, the logs would not be generated.
As this is a private provider-managed network, however, these two problems are not
severe. To guarantee the expected high quality of experience for its customers Telefonica
rigorously controls the load of its network elements. Hence, the probability of these
events occurring is low. Such low probabilities, put together with the scale of the
dataset, gives guarantees that possible errors are rare and therefore have negligible
impact on statistics. Finally, as this is a data trace, i.e., a timzB“tvmpzy ordered record
of all requests of the IPTV system, it is very important that the network elements
are synchronised. In Telefonica’s IPTV network all network elements are synchronised
using the Network Time Protocol (NTP). The time reference is taken from a reference
clock in Telefonica’s network. NTP is known to achieve (worldwide) accuracy in the
range of 1 to 50ms [190], which is one order of magnitude better than what is necessary
for the current study.
ICGCG Dvtv xhvrvxterisvtion
The trace includes all channel switching events from April 16th 2007 to October 20th
2007, six months in total. The log scales up to 150 TV channels, 680 DSLAMs, and
255 thousand users. Table 4.1 summarises these statistics. These data do not include
1Xron i“ v timzBwv“zy jow “xhzyulzr in jnixBlikz xomputzr opzrvting “y“tzm“C Xron znvwlz“ u“zr“
to “xhzyulz jow“ to run pzrioyixvlly vt xzrtvin timz“ or yvtz“C Xron i“ yrivzn wy v xrontvw filzA v
xonfigurvtion filz thvt “pzxifiz“ “hzll xommvny“ to run pzrioyixvlly on v givzn “xhzyulzC
2\Tfc magl nspnmsc md Imaeclgm gs npmtgdgle tgdcm scptgac rm assrmmcps, parfcp rfal amllcargle
lmes"A I fiv“ tolyC
ICG Dvtvset 5H
core IP network
IPTV
head-end
core
network
regional
network
metro
network
TV
channels
STB
Centralised log
collection server
Local area
server
send_log
message
send_igmp_join/leave
message
IGMP join/leave
message
access
network
Figure 4.2: Data collection process
any other information. For example, they do not capture performance related metrics
such as network latency, jitter, and loss of the IPTV streams. They also do not capture
the remote control commands issued by the user to switch channels.
A single line from the trace has the information on channel switching presented
with the following format:
9JLNTE> 9A>V> 9ELRO>79JFN>79PBC> [9>>.9B>.9C>.9A>.9B>.9C>] 9PEBIC.>
,9PILT.>7 Jritfcapt cifekt 9FM\JRI> ifkh RM colj 9FM\POC>,
mlot 9PEBIC>,9PILT>,9MLOT>, smf,scf 9SMF>,9SCF>Y$
For the purposes of this study, the most relevant information present in each line is
5I bethoyology vny yvtvset
Table 4.1: Dataset statistics
irvxz yurvtion 6 months
cumwzr of u“zr“ 255 thousand
cumwzr of DhaVb“ 680
Vvzrvgz numwzr of yvily xhvnnzl “fiitxhing zvznt“ 13 million
hizz of thz yvtv“zt 700 GB
the following:
1. 9JLNTE> 9A>V> 9ELRO>79JFN>79PBC>
Timestamp in units of seconds.
2. 9>>.9B>.9C>.9A>
IP address of the DSLAM.
3. 9FM\JRI>
IP address of the multicast group (in most cases, a TV channel).
4. 9FM\POC>
IP address of the source (in most cases, a Set Top Box).
5. ifkh RM
Multicast option of joining (UP) or leaving (DOWN) a channel.
An example line of the log follows.
>mo .3 --7-27-0 [.4/./1./.2...0./21] .,57 Jritfcapt cifekt
/06.-.-..1 ifkh RM colj FM .-.6-...41, mlot .,/,.5, smf,scf 5,02
ICGCH Dvtv xlevning vny pvrsing
The dataset made available by Telefonica has some information that is irrelevant for the
purposes of this dissertation, which means it can be filtered without loss. In addition,
the original format of the IPTV traces is not the most convenient for the study I present
in Chapters 5, 6, and 7. For this reason I decided to clean and parse the data, a process
I describe in this section.
The first step of the process is to filter all irrelevant information. This includes
removing messages from all devices other than the Set Top Boxes, all messages from
non-TV channels, and also some outliers. In more detail:
ICG Dvtvset 55
1. All messages with a source IP address outside the .-.u.u.u range are removed.
.-.u.u.u is the STB IP address format. However, there is a very small percentage
(less than 0.1% of the total) of requests with a different source IP address. An
example is the address -.-.-.-, the default IP address of every STB. The first
time an STB is plugged on the network, it starts sending packets with this address
until the automatic installer assigns it the correct IP address.
2. All messages with a multicast IP address outside the /06.-.u.u range are re-
moved. These are non-TV multicast groups. There is also a very small percentage
of requests in this category (again, less than 0.1% of the total). These include
multicast groups used to manage Set Top Boxes, for example for bootstrapping
and for upgrading the software.
3. Some outliers are also removed. When I analysed the data I detected a strange
behaviour from some STBs. In particular, some STBs sent IGMP signals with a
fixed periodicity for the trace duration (the whole 6 months). Such non-human
behaviour is usually a characteristic of devices included in the network for testing
purposes. I assume this is the case, and filter the information from these STBs.
Note that I found only 4 such devices (out of the 255 thousand), hence its removal
or inclusion would always be statistically irrelevant.
The second step of the process is to parse the data for a more convenient format.
As explained in the previous subsection, each line of the trace includes information —
namely, the STB IP address — which allows the separation of the channel switching
events from each STB. I therefore create a single file per STB that includes all channel
switch requests made by the users of a specific household. After this final step, the
cleaned and parsed channel switching log used in this study has the following format:
#cfopt ifke lkiv7
PTB FM : 9PTB\FM\>AAOBPP> API>J FM : 9API>J\FM>
#colj tee peclka tl tee cfkai ifke7
9JLNTE> 9A>V> 9ELRO>79JFNRTBP>79PBCLNAP>y9TVMB>y9CE>NNBI\NO>
#...
Each log file includes, in the first line, the STB IP address and the DSLAM IP
address. The rest are all channel switching requests — one per line — made by that
STB during the whole period of the trace. Each of these lines includes the timestamp
(date and time with the precision of one second), the request type (RM or ALTN), and
the channel number. An example of part of one such parsed file follows.
56 bethoyology vny yvtvset
PTB FM : .-.41.26.65 API>J FM : .4/./1./1-..
Jri . --71.703yRMy/0
Jri . --71.711yALTNy/0
Jri . --71.711yRMy/2
Jri . --71.715yALTNy/2
Jri . --71.715yRMy.5/
Jri . --71.72.yALTNy.5/
Jri . --71.72.yRMy0-
Jri . --71.72.yALTNy0-
Jri . --71370/yRMy0.
ICGCI kvliyvtion of the yvtvset
Trace-driven analysis have several advantages when compared with other evaluation
methods. These include, but are not limited to [181], the similarity with the actual
implementation of the system under evaluation and the fact of a trace being an accurate
workload offering high level of detail. But using a data-trace from a real system has
nonetheless some disadvantages. One issue that has to be taken into account is that
of representativeness. Traces taken from one system may not be representative of the
workload on another system. For this reason, validating the dataset is an important
part of the process, to help ensure that the algorithms and ideas one wants to evaluate
are done so on correct, useful, and representative data.
To validate a data trace, a possibility is to obtain a different trace under a different
environment and use it to validate the original dataset [113]. An alternative is to
compare the results obtained from the analysis of the same type of system using a
different dataset, with the same analysis using our own. This is the technique I use in
this section in order to validate the dataset used in the current study. As explained
in Chapter 3, Qiu et al [160, 161] have analysed an IPTV system using a different
dataset from the one used in the current study1. These researchers measured several
characteristics of their IPTV system, explored TV channel popularity and characterised
and modelled user activities. By comparing a relevant subset of their results with
those obtained by analysing Telefonica’s dataset, I believe it is possible to validate the
correctness, usefulness and representativeness of these traces.
The first characteristic of an IPTV system I analyse is the number of online users
during the course of a representative week. In [160] and [161] Qiu zt vlC found very
1ihz vuthor“ vnvly“zy v yvtv“zt from Vi&iC
ICG Dvtvset 5L
strong diurnal patterns, with daily peaks at around 9PM, followed by a quick decrease
in the number of online Set Top Boxes, reaching a daily minimum at 4AM, and then
steadily rumping up during the course of the day. Unfortunately, Telefonica’s trace does
not include information on when an STB is turned off. For this reason I differentiate
an online from an offline user by assuming a user is offline when the xhvnnzl yfizll
timz, i.e., the time an STB stays tuned in a channel, is above one hour. This is the
same procedure followed by Cha zt vlC [42] for the same purpose. I am thus assuming
that users leave the STB on even when they switch the TV off. Of course, some users
may watch TV for more than one hour without switching channels, especially when
watching movie channels. In Figure 4.3 I show the time series of the number of online
users over a representative trace period. The results are very similar to Qiu’s. The
strong diurnal pattern is present, with a daily peak in the evening (at around 10PM
in Spain). It is also interesting to note that the lowest evening peaks correspond to
Friday and Saturdays.
20000
40000
60000
0 1 2 3 4 5 6 7
Days
N
u
m
b
er
 o
f 
o
n
li
n
e 
u
se
rs
Figure 4.3: Number of viewers during a representative week
In [161] Qiu zt vlC examined the long term distribution of channel popularity us-
ing both xhvnnzl vxxz““ frzquznxy — the number of channel switching requests to the
channel — and channel dwell time (defined above). They observed that both distribu-
tions are very similar (I return to this similarity later), exhibiting high skewness, with
the top 10% of channels accounting more than 90% of channel accesses. By analysing
the channel switching events of a subset of 2200 random users (over the entire 6 month
5M bethoyology vny yvtvset
period), I also observed the same trend in my dataset, as Figure 4.4 attests (data points
labelled “empirical data”). This figure shows the channel access frequency as a function
of channel ranking. A channel rank of 1 indicates the most popular channel and the
unpopular channels are at the tail of the distribution.
1
100
10000
● ● ● ●
● ●
● ● ● ● ● ● ● ●
● ● ● ● ●
●●●●●●●●
●●●●●●●●●●●●●●●●
●
●●●●
●●
●●●
●
●●●●●●●●●●●●●●●
●
●●
●
●●●●●●●●
●●●
●●
●
●
●
●
●●
●
●●●●●
●
●●
●
●
●
●●●
●
●
●●
●
●
●
●
●
●
●
● ● ● ● ● ● ● ●
●
●
●●●●●●●●●●●●●●●●●●
●●●●●●●
●●●●●●
●●●●●●●●●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
Mixed model
Empirical data
10 50 90 130
Channel index sorted by popularity (log scale)
A
cc
es
s 
fr
eq
u
en
cy
 (
lo
g
 s
ca
le
)
Figure 4.4: Channel popularity distribution
The high skewness of popularity is usually modelled using Zipf-like distributions.
Qiu zt vlC have indeed shown that the 10% most popular channels can be modelled
with this type of distribution. However, the exponential function achieves a better fit
for the large “body” part of the distribution function. They thus proposed a hybrid
model with the probability density function expressed as follows,
fo (i) =
{
C1i
−RC0 i < 10% of available channels,
e−+X2RC0 others,
(4.1)
The parameters Qiu zt vlC found for the Zipf-like distribution, f1(i) = C1i
−, and
for the exponential distribution, e−+X2 , are presented in Table 4.2. Note that C0 is a
normalisation factor such that f0(·) is a well-defined probability density function. This
popularity model also fits quite well with the Telefonica dataset. Again, I analysed the
channel switching events of a subset of 2200 random users (over the entire 6 month
period). I fixed the values for C1 and C2, and the values for  and  that fit the
empirical data are also presented in Table 4.2. The result can be observed in Figure 4.4
(data points labelled “Mixed model”). The lower value of  in the Zipf-like part of the
ICG Dvtvset 5N
Table 4.2: Parameters of the channel popularity models
AT&T trace Telefonica trace
 0.513 0.2
C1 12.642 12.642
 0.006 0.05
C2 2.392 2.392
distribution means that the popularity of the top 10% channels offered by Telefonica
decays slower that in AT&T’s case. This may be justified by the size of the population.
AT&T’s dataset has four times more users than Telefonica’s and hence the difference
in popularity between top channels may be more pronounced. This is just a conjecture,
however. The justification for the higher  parameter from the exponential distribution
may be easier to accept. The higher  value for Telefonica’s dataset means the popu-
larity of the bottom 90% channels decays more rapidly than in AT&T’s. The reason
may be the total number of channels. In AT&T’s network, 630 channels compose the
bottom 90% list of channels, which is five times Telefonica’s figure. This may justify
the higher skewness of Telefonica’s graph, as users have less channel options.
I mentioned before that channel popularity based on dwell time and channel pop-
ularity based on access frequency produce similar results, as reported by Qiu zt vlC In
their paper [161] the authors indeed found a very strong correlation between these two
popularity measures. I found the same correlation in Telefonica’s dataset. This can be
observed in Figure 4.5. This figure shows the scatter plot of the ranks of the channels
according to each popularity measure. The fl -axis shows the popularity rank according
to channel access frequency, while the rank according to channel dwell time is shown
on the y-axis. The points are spread well along the diagonal line, indicating strong
correlation. Their Spearman rank correlation coefficient and their Pearson correlation
coefficient are both equal to 0.97, demonstrating the strong correlation. Very similarly,
Qiu zt vlC reported the values 0.98 and 0.97 for these coefficients, respectively.
An insight that is important to gain for the current work (in particular to Chapter
5) from the data is to understand how IPTV users switch channels. Do users switch
linzvrly, up or down to the next or previous TV channel, or do they perform more
targeted switching, with the user switching intentionally to a specific channel of choice
(thus “jumping” several channels)? By analysing the whole dataset from Telefonica I
observed that 55% of all channel switching was linear. Qiu zt vlC [160] reported 56% in
the AT&T dataset. From these, in Telefonica’s network 69% are up-channel-switches.
This figure is equal to 72% in AT&T’s case.
6E bethoyology vny yvtvset
20
40
60
80
100
120
0 20 40 60 80 100 120
Ranking based on channel access frequency
R
an
k
in
g
 b
as
ed
 o
n
 c
h
an
n
el
 d
w
el
l 
ti
m
e
Figure 4.5: Correlation between channel access frequency and channel dwell time
There is no such thing as a validated dataset [113], but I believe the similarity of
the results obtained from this analysis of Telefonica’s dataset with that from the AT&T
studies helps increase the degree of confidence in the dataset used and in the results I
present in this dissertation.
Chvpter 5
geyuxing xhvnnel xhvnge yelvy
One of the major concerns of IPTV network deployment is channel change delay (also
known as zapping delay). As explained in Chapter 2 (Section 2.1.1), synchronisation
and buffering of media streams can cause channel change delays of several seconds.
The main concern in the industry and in the research community has been, in fact, to
try to improve the performance on these two aspects, and several solutions have been
proposed. One such solution is preyixtive preBjoining of ik xhvnnels. In this
scheme each Set Top Box (STB) simultaneously joins additional multicast groups (TV
channels) along with the one that is requested by the user. If the user switches to any of
these channels next, the switching latency is virtually eliminated, and user experience
is improved. The negative impact of this solution is additional load in the access
network, and the evaluation presented in this chapter looks at the tradeoff between
pre-join advantage in reduced switching latency versus the access network bandwidth
cost.
As observed from the analysis of the data traces presented in Chapter 4 (Section
4.2), most channel switching events are relatively predictable: users very frequently
switch linearly, up or down to the next TV channel. This favours this specific type of
solution to the channel change delay problem. Previous work on this subject [81, 189]
used simple mathematical models to perform analytical studies or to generate synthetic
data traces to evaluate these pre-joining methods. I showed in Chapter 4 (Section
4.1) that these models are conservative in terms of the number of channel switches
a user performs during zapping periods. They therefore do not demonstrate the true
potential of predictive pre-joining solutions. This is an important motivation to perform
an empirical analysis using the IPTV dataset available. Such realistic trace-driven
analysis is the main differentiating point of my contribution. This is, to the best of my
6G geyuxing xhvnnel xhvnge yelvy
knowledge, the first empirical study of channel change delay reduction techniques.
The first pre-joining scheme I analyse in this chapter is very simple. In this scheme
the neighbouring channels (i.e., the channels adjacent to the requested one) are pre-
joined by the Set Top Box alongside the requested channel, during zapping periods.
Notwithstanding the simplicity of the scheme, the trace-driven analysis shows that
the zapping delay can be virtually eliminated for a significant percentage of channel
switching requests. For example, when sending the previous and the next channel con-
currently with the requested one, for only one minute after a zapping event, switching
delay is eliminated for near half of all channel switching requests. Importantly, this
result is achieved with a negligible increase of bandwidth utilisation in the access link.
Two other schemes are evaluated. The first considers pre-joining popular TV channels,
but the results are unsatisfactory. The second is a personalised scheme where user
behaviour is tracked to decide which channels to pre-join next. The improvement of
this scheme over the simpler version is also insignificant.
5CF Introyuxtion
The offer of TV services over IP networks is very attractive as it represents a new source
of revenues for network operators. IPTV offers network providers greater flexibility,
while at the same time offering users a whole new range of applications. In order to
compete in this market, IPTV operators have to at least guarantee the same Quality of
Experience (QoE) offered by cable networks or over the air broadcasts. In this respect,
one of the major concerns of IPTV network deployment is channel change delay (also
known as zapping delay). This is the delay between the time the user switches to a
particular TV channel and the time when the content is displayed on the TV screen.
An analysis to the causes of this delay was presented in Chapter 2 (Section 2.1.1). I
refer the reader to Figure 2.2 in particular. When a user switches to a new TV channel
using his or her remote control, the STB issues a new channel request towards the
network. After a certain time (the nztfiork yzlvy), the first packets of that particular
multicast group start flowing. Before play-out the STB still has to “ynxhroni“z with
the video stream (it has to wait for the next I-frame) and wuffzr some packets (to avoid
starvation and to compensate for networked-introduced jitter and packet-reordering
delay). The whole process therefore takes some time. Synchronisation and buffering of
media streams can cause channel change delays of several seconds [80, 176, 182, 189].
Figure 2.3 in Chapter 2 pictorially summarises the contribution of each component of
channel change time. It is known that this figure should be below 430ms to guarantee an
5CF Introyuxtion 6H
acceptable user experience [128], so this is a major concern for IPTV service providers
that want to compete in this market.
By analysing the dataset described in Chapter 4, I observe that most channel switch-
ing events are linear: users switch up or down to the next TV channel very frequently1.
Also, even when zapping is not linear, the “jump distance”2 is usually small (i.e., there
is a high probability for the user to switch to one of the neighbouring channels). These
facts can be observed in Figure 5.1. This figure presents the Cumulative Distribu-
tion Function of the “jump distance” considering the analysis of the whole dataset
(255 thousand users, 6 months, 13 million channel switches per day on average). The
probability of zapping linearly (“jump distance” equal to 1) is close to 55%, and the
probability of jumping to a close neighbour is also very high. For example, 80% of
all channel change requests are to channels not more than six channels apart (“jump
distance” equal to 6).
0%
20%
40%
60%
80%
100%
55% of all channel switching 
is to the next or previous channel
80% of all channel switching 
is to channels not more than 6 apart
0 1 2 3 4 5 6 7 8 9 10
Jump distance
C
D
F
 o
f 
ju
m
p
 d
is
ta
n
ce
Figure 5.1: Cumulative distribution of zapping jump distance
This kind of user behaviour is very favourable for a specific type of solution to
the channel change delay problem, namely, predictive pre-joining of TV channels. As
explained in Chapter 3 (Section 3.2.4), in these schemes each Set Top Box (STB)
simultaneously joins additional multicast groups along with the one that is requested by
the user, thus anticipating future user behaviour. These schemes are thus based on the
1It i“ rzlzvvnt to mzntion thvt vt thz timz of yvtv xollzxtion thz yzployzy Ieik “y“tzm “upportzy
vn Elzxtronix erogrvm GuiyzC
2V““uming thvt thz ik xhvnnzl“ vrz numwzrzy v“ v “ortzy li“tA v“ i“ xommonA thz qjump yi“tvnxz7
i“ thz yiffzrznxz wztfizzn thz numwzr of thz xhvnnzl “fiitxhzy to vny thz numwzr of thz xhvnnzl “fiitxhzy
fromC
6I geyuxing xhvnnel xhvnge yelvy
prediction of the next TV channel the user will switch to. If the prediction is right, the
user will experience a small zapping delay because the channel is already synchronised
in the STB. The negative impact of this solution is additional load in the access network,
so there is a tradeoff between the advantage in reduced switching latency versus the
access network bandwidth cost. Previous work evaluated this technique using analytical
techniques or simulations based on simple mathematical models. As was proved by Qiu
zt vlC [160] and as I demonstrated in Chapter 4 (Section 4.1), these simple models do
not capture IPTV user behaviour well. For this reason, in this chapter, I perform
a trace-driven analysis using the Telefonica dataset to evaluate this solution to the
channel change delay problem. To the best of my knowledge, this is the first empirical
study where channel change delay reduction techniques are evaluated using real IPTV
usage data from an operational network provider. That is the main contribution of this
work.
I consider several pre-joining schemes. In the first, the set of channels pre-joined
are the neighbouring channels (i.e., channels adjacent to the requested one). These
neighbouring channels are synchronised and buffered together with the requested one.
Therefore, if the user decides to switch to any of these channels, the switching delay
experienced is virtually zero. These additional channels are not sent to the STB contin-
uously; they are kept during zapping periods only, to assure the scheme is bandwidth
efficient1. One of the main advantages of this scheme is its simplicity. Notwithstand-
ing its lack of sophistication, the trace-driven analysis shows that the zapping delay
can be virtually eliminated for a significant percentage of channel switching requests.
For example, when sending only two channels concurrently (the previous and the next,
respectively, thus assuming that the user will zap linearly), for only one minute after
a zapping event, switching delay is eliminated for around 45% of all channel switch-
ing requests. This figure jumps to 60% if one considers zapping periods only (periods
when the user is surfing/browsing, i.e., actively switching between channels). If the
access network has enough bandwidth available to increase the number of neighbouring
channels to eight, around 80% of all switching requests during zapping periods will
experience no delay.
Globally, this scheme offers very interesting results. I demonstrate in this chapter
that this simple scheme has a performance close to that of an optimal predictor. How-
ever, I also observe that user behaviour can vary significantly: it is true that many
users enjoy zapping up and down, but others seem to zap less linearly. Therefore, while
1Wvnyfiiyth inzffixiznxy fiv“ onz of thz prowlzm“ of thz originvl pvpzr propo“ing przyixtivz przB
joining of ik xhvnnzl“ pJFrA v“ zxplvinzy in Xhvptzr H =hzxtion HCGCI)C
5CF Introyuxtion 65
some users would benefit hugely from using this scheme, others would see only a rela-
tively small improvement. With this limitation in mind I also consider other schemes
to see if user experience can be improved for a wider audience. I first test a scheme
where the most popular channels are pre-joined, either alone or together with some
neighbours. This scheme proves inefficient when compared with pre-joining neighbours
only. I also evaluate a personalised scheme. In this scheme user behaviour is tracked
by maintaining information on user actions: does the user have favourite channels,
preferring to switch to a particular channel or set of channels, or does he/she prefer to
zap linearly? This scheme also accommodates temporal dynamics, capturing changes
in user behaviour over time. In the end, the added complexity of the scheme does not
result in an improvement over the simple scheme of pre-joining neighbours only.
As explained in more detail in Chapter 2 (Section 2.3), current operational IPTV
networks are “walled gardens”, with all TV channels distributed to the edge of the
network (to the DSLAMs). However, due to access link bandwidth limitations, only
one or two TV channels are distributed from this edge point to the Set Top Box. It
is therefore important to underline that I assume in this study the access network is
able to accommodate the peak bandwidth needed to distribute several TV channels
concurrently. Most systems today distribute Standard Definition (SD) TV channels
using MPEG-2, requiring 4 Mbps guaranteed bit rate per channel, thus sending channels
in parallel increases the bandwidth requirements proportionally. I believe, however,
that this is not a serious limiting factor of the type of schemes I analyse. In fact, in
most OECD countries access networks already offer tens of Mbps of average download
speeds. Japan and South Korea, for instance, have an average broadband speed close
to 100 Mbps, with Japan already offering 1Gbps to some users, and it is expected other
countries to follow this trend in the near future [157]. More importantly, by using
these schemes the increase of bandwidth utilisation in the access link is negligible,
since the concurrent channels are distributed to the STB during zapping periods only.
I also assume that the STB is able to process (i.e., synchronise and buffer) several TV
channels in parallel. In this study I take both these limitations in consideration, and
restrict the number of neighbouring TV channels sent in parallel.
The rest of the chapter is organised as follows. The first scheme evaluated — pre-
joining neighbouring channels — is described in Section 5.2. In section 5.3 I detail
the methodology used to evaluate this (and the other) scheme(s), and in section 5.4 I
present the results for this simple technique. The two sections that follow present and
evaluate the other schemes. First I consider pre-joining the most popular channels.
Then I propose and present results of a personalised scheme where user behaviour is
66 geyuxing xhvnnel xhvnge yelvy
tracked. In Section 5.7 I discuss the advantages and disadvantages of the schemes under
analysis, and I conclude this chapter in Section 5.8.
5CG himple sxhemeO preBjoining neighwouring xhvnnels
In current IPTV systems, when a user requests a TV channel using the remote control,
this single channel is requested from the network and delivered to the Set Top Box. The
idea behind all the schemes evaluated in this chapter is to pre-join, together with the
channel requested, an additional set of TV channels concurrently, based on some sort of
prediction. These will be synchronised and buffered simultaneously with the requested
one. Therefore, if the following request is for a channel already present in the STB,
there is no network, synchronisation, or buffering delay, and switching delay will be the
result of STB processing only, thus virtually zero (i.e., way below the 430ms needed to
guarantee an acceptable viewing experience). The neighbouring channels stay in the
STB for a limited, predefined period. I call this period the xonxurrent xhvnnel time.
After this time, the STB sends IGMP leave requests and the neighbouring channels are
removed.
The first scheme considered is very simple: the extra channels to pre-join are the
neighbouring channels (i.e., channels adjacent to the requested one), and they are
distributed to the STB during zapping periods only. An example of the use of this
method is shown in Figure 5.2. In the figure I assume that after a channel switching
event only two neighbouring channels (previous and next) are pre-joined additionally
to the requested one. I also assume the user is in viewing mode in the beginning, i.e., he
or she is settled watching channel x. The user then switches to channel y. Right after
the channel change the STB enters in “zapping mode” and requests three TV channels
from the network: the channel the user switched to, y, the next channel, y + 1, and
the previous channel, y − 1. As there is no video data for channel y in the STB before
the change, the viewer has to wait for the synchronisation and buffering of the video
streams, thus experiencing zapping delay (represented by the gray box). When the user
switches again, this time to channel y+1, he or she experiences virtually no delay, since
the channel is already being received by the STB. As the user is in up-channel-switching
mode, the STB sends an IGMP leave message from channel y − 1, and a join message
for channel y + 2. When the user exits zapping mode1, i.e., when it settles in channel
y+1, the STB leaves the neighbouring channels, y and y+2. In the following channel
switching event the user will therefore experience zapping delay, independently of the
1Vftzr thz xhvnnzl xonxurrznt timz zlvp“ingC
5CH bethoyology 6L
channel switched to.
channel y
channel y+1
channel y-1
channel x
Switch to
channel y
Switch to
channel y+1
viewing mode zapping mode viewing mode
synchronisation and
buffering delay
channel concurrent time
channel y+2
t
Figure 5.2: Predictive pre-joining of TV channels
5CH bethoyology
The schemes proposed in this chapter are evaluated by means of a trace-driven analysis
for the reasons explained in Chapter 4 (Section 4.1). The IPTV trace detailed in that
chapter is used as input to the analysis performed.
A small detail needs clarification beforehand. Sometimes users zap linearly, from
one channel to the next, swiftly (in less than the normvl Ieik “fiitxhing yzlvy, which
I consider to be 2 seconds in the rest of this chapter1). In these cases, the STB may
not have time to synchronise to the requested channel, or to any of its neighbours.
However, all these channels are already in synchronisation mode. Therefore, if the next
change is to a channel already in the STB, the switching delay, albeit not being zero,
will be less than the normal delay. In this case, I say the user experienced pvrtivl yzlvy
(some value between zero and the normvl yzlvy).
To evaluate the pre-joining schemes I developed a Python script that checks each
line of the input trace to obtain each switching event. The current switching event is
then compared with the previous, and one of these actions is performed:
1. If the time between two user switching events is above the concurrent channel
time, no additional channel is in the STB, and therefore the user experiences the
normvl “fiitxhing yzlvy. The counter klojai\aeiav is incremented.
1ihz rzv“on fihy I xon“iyzr G “zxony“ i“ zxplvinzy in Xhvptzr G =hzxtion GCFCF)C ihi“ i“ vl“o thz
vvluz u“uvlly xon“iyzrzy in othzr “tuyiz“ on przyixtivz przBjoiningA “uxh v“ pFMNrA for in“tvnxzC
6M geyuxing xhvnnel xhvnge yelvy
2. If the time between two user switching events is below the concurrent channel
time, there are additional channels in the STB. So, one of these three situations
occurs:
v If the user switches to a different channel from the ones in the STB, it will
experience the normvl yzlvy. The counter klojai\aeiav is incremented.
w If the user switches to one of the neighbours in the STB, and if the time between
two user switching events is above or equal to 2 seconds, the user experiences
virtually no delay. This is due to the fact that the channel is in the STB,
and is already synchronised. The counter kl\aeiav is incremented.
x If the user switches to one of the neighbours requested by the STB, but the time
between two user switching events is under 2 seconds, the user experiences
pvrtivl yzlvy. As explained above, this is due to the fact that the channel
was already requested by the STB, but the user zapped rapidly, so it did
not have time to synchronise. The counter maotfai\aeiav is incremented.
Figure 5.3 illustrates the proposed methodology with a simple example1. When the
user turns the STB on it switches to channel 23 at 12:41:36am. This is translated into
three IGMP join messages, to channels 22, 23, and 24, respectively, sent to the network.
After a couple of seconds (the normal switching delay considered) these three TV chan-
nels are being distributed to the STB, so at the time of the next switching event, at
12:41:44am, they are all being received simultaneously. The user then switches to chan-
nel number 24. The channel is being received by the STB, already synchronised, so the
user will experience virtually no channel change delay (the counter kl\aeiav is incre-
mented). For that reason, only a join message is sent to channel 252 (channels 23 and
24 are already being received). Next, the user switches to channel 182. As this channel
is not available in the STB, the STB has to send a join message to channels 181 to 183,
and the user will experience the normal zapping delay (the counter klojai\aeiav is
incremented). In any case, after the concurrent channel time two IGMP leave messages
are sent by the STB to leave the additional channels. For this reason, and assuming
the concurrent channel time is equal to one minute, although at 12:46:32am the user
performs up-channel-switching to channel 31, this channel is not being distributed to
the STB anymore, and so the user will experience the normal channel switching delay.
Again, the counter klojai\aeiav is incremented.
1ihi“ figurz i“ wv“zy on thz rzfzrznxz vrxhitzxturz prz“zntzy in [igurz GCJC
2I omit IGbe lzvvz mz““vgz“C
5CI Zvvluvtion 6N
STB
access network
STB IP = 10.74.59.98 DSLAM IP = 172.24.240.1
Jul 1 00:41:36|23
Jul 1 00:41:44|24
Jul 1 00:41:48|182
Jul 1 00:41:51|30
Jul 1 00:46:32|31
Parsed log file
igmp_join(22,23,24)
igmp_join(25)
igmp_join(181,182,183)
igmp_join(29,30,31)
igmp_join(30,31,32)
Join messages
1, 2, 3, 4, 5
input
1.
2.
3.
4.
5.
Join messages
{}
{22,23,24}
{23,24,25}
{181,182,183}
{30}
TV channels in the STB
prior to each step:
1.
2.
3.
4.
5.
1.
2.
3.
4.
5.
Figure 5.3: Proposed methodology
5CI Evvluvtion
I consider two dimensions in the analysis. The first is the time the neighbouring chan-
nels are sent concurrently to the STB, the xonxurrent xhvnnel time, already referred
to above. After this time, the STB sends IGMP leave requests and the neighbouring
channels stop being distributed. The scheme leading to the best results would be to
send these channels vlfivy“, i.e., never leaving the neighbouring channels. However, this
is inefficient, as it unnecessarily increases access link bandwidth utilisation. Therefore,
in the schemes under analysis I maintain the neighbouring channels in the STB during
a small period, between 10 seconds and 2 minutes. But I also present the results for
the case “always” referred to before, for comparison. I choose this range of values in
accordance with the definition of zapping (or surfing) periods in previous research [42],
which is also in line with the way Nielsen Media Research demarcates viewing events.
The second dimension is the numwer of neighwouring xhvnnels to send concur-
LE geyuxing xhvnnel xhvnge yelvy
rently. Current Set Top Boxes typically receive one or two TV channels in parallel.
Although the technology for a STB to stream more channels in parallel is available
today, these are typically low cost devices, thus constrained in terms of its processing
and memory capabilities. As the costs of processors, memory and storage continue
to fall, devices capable of processing more channels in parallel will plausibly become
cost-effective. Anyway, considering the limitations of STB processing and of access link
bandwidth, I decide to restrict the number of concurrent channels to a minimum of 2
and a maximum of 8. In fact, the gain of sending more channels would be small, as
can be inferred from Figure 5.1. All the results presented in this section and in the
rest of the chapter arise from the analysis of the whole data set (255 thousand users, 6
months, 13 million channel switches per day on average).
Figure 5.4 illustrates the percentage of switching events that experience virtually
zero delay. The x-axis is the percentage of switching requests that experience no delay,
and the y-axis is the concurrent channel time, i . The main conclusion is that by
using this very simple scheme we can reduce zapping delay to a significant number of
switching events. For example, by pre-joining only 2 neighbours, the previous and the
next channel, for only one minute, the delay is reduced to virtually zero to around 45%
of the switching events. If the access link has enough bandwidth available to increase
the number of neighbouring channels to eight, around 60% of all switching requests
experience no delay. It is important to underline that when a user watches a long
program, without switching channels for an extended period of time, then vny scheme
except the one that always joins the predicted channels will have a delay. As a final
note, only 2 to 3% of the requests will experience partial delay in all cases. For this
reason, and to keep the presentation of the results as clear as possible, I do not include
this information in the figures.
T=Always
T=120 sec
T=60 sec
T=30 sec
T=10 sec ●
●
●
●
●
45% of all requests
60% of all requests
25% 50% 75% 100%
Percentage of switching requests
Neighbours ● 2 4 6 8 Optimal predictor
Figure 5.4: Percentage of requests that experience no delay by using the simple scheme,
for various values of channel concurrent time i and number of neighbours
5CI Zvvluvtion LF
In Figure 5.5 I consider only zapping periods. Focusing on zapping periods in
detail is important, because arguably it is in these periods that the user expects a
swift zapping experience. I consider that a user is in “zapping mode” if the time
between consecutive switching requests is less than one minute (again, in accordance
with previous research [42] and Nielsen Media Research [147]). Therefore, all events
for which the switching time was above one minute were removed (for this reason, I
logically do not include the results with xonxurrznt xhvnnzl timz above one minute).
One can see that, for example, by pre-joining only 4 neighbours for one minute, more
than 70% of the switching requests during zapping periods will experience virtually no
delay.
T=60 sec
T=30 sec
T=10 sec ●
●
●
70% of all requests
25% 50% 75% 100%
Percentage of switching requests
Neighbours ● 2 4 6 8 Optimal predictor
Figure 5.5: Percentage of requests that experience no delay by using the simple scheme
during zapping periods only, for various values of channel concurrent time i and number
of neighbours
The pre-joining schemes that are the subject of this study involve the prediction
of the next channel an IPTV user is going to switch to. Obviously, sometimes these
predictions are wrong, so it is important to understand how these schemes compare
with an algorithm based on complete knowledge. Such a predictor always knows to
which TV channel the user will switch next, and is therefore used as a benchmark for
comparison. For convenience, I refer to this predictor as “optimal”. Such “optimal”
predictor would not do much better than most of the schemes I tested, as can be
attested from the previous figures1, where I already included its results. To make this
comparison clearer, in Figure 5.6, I show the pzrformvnxz gvp of each of the schemes
under evaluation to the optimal predictor. This performance gap is defined as the
difference between the percentage of requests the optimal predictor would benefit and
the same percentage using the simple scheme. It is interesting to note that the simple
1It i“ fiorth noting thz =vt v fir“t glvnxz) xuriou“ fvxt of thz optimvl przyixtor not vxhizving FEE:
zvzn fiith T R lwwlysC ihz rzv“on i“ thz G to H: of thz rzquz“t“ thvt zxpzriznxz pvrtivl yzlvyC
LG geyuxing xhvnnel xhvnge yelvy
scheme produces results that are not very distant from the optimal. An optimal scheme
would not perform much better than a scheme that sends 6 or 8 neighbours, for instance.
T=Always
T=120 sec
T=60 sec
T=30 sec
T=10 sec ●
●
●
●
●
0% 10% 20% 30% 40% 50%
Performance gap (percentage)
Neighbours ● 2 4 6 8
Figure 5.6: Performance gap between optimal predictor and the simple scheme for
various values of channel concurrent time i and number of neighbours
Currently, most IPTV service providers distribute TV content in SD format encap-
sulated as MPEG-2 streams, so each TV channel needs 4 Mbps of guaranteed bitrate.
As explained before, it is important that the access link can accommodate the distri-
bution of several TV channels concurrently. I assume that is the case. However, when
several TV channels are sent in the access link other broadband applications (P2P,
web browsing, etc.) are affected, so it is important to quantify its impact. That is the
purpose of Figure 5.7. In this figure I illustrate the average bandwidth consumption
across the trace period, to understand the impact these simple schemes will have in this
particular. One can observe that by limiting the xonxurrznt xhvnnzl timz to zapping
periods only, the average bandwidth is very close to the 4 Mbps current IPTV services
usually require. This is due to the fact that zapping periods are relatively rare events
during the course of a normal day [166]. So, even though the access link will have
to accommodate peaks of high bandwidth consumption1 during zapping periods, these
average out during the course of the day.
5C5 ereBjoining populvr ik xhvnnels
Some measurement studies on IPTV analysed channel popularity in detail, concluding
that TV channels popularity is highly skewed and can be characterised by a Zipf-like
distribution for top channels and an exponential distribution for non-popular ones [42,
161]. An interesting question is thus to investigate if pre-joining the most popular
1cotz thvt thz plot T R lwwlys in thz figurz vl“o rzprz“znt“ thz“z pzvk“C
5C5 ereBjoining populvr ik xhvnnels LH
Optimal predictor
8 neighbours
6 neighbours
4 neighbours
2 neighbours
Optimal predictor
8 neighbours
6 neighbours
4 neighbours
2 neighbours
Optimal predictor
8 neighbours
6 neighbours
4 neighbours
2 neighbours
Optimal predictor
8 neighbours
6 neighbours
4 neighbours
2 neighbours
Optimal predictor
8 neighbours
6 neighbours
4 neighbours
2 neighbours
Average bandwidth close to 4 Mbps
T
=
1
0
 sec
T
=
3
0
 sec
T
=
6
0
 sec
T
=
1
2
0
 sec
T
=
A
lw
ay
s
0 5 10 15 20 25 30 35
Average bandwidth consumption (Mbps)
Figure 5.7: Average bandwidth consumed by the simple scheme for various values
of channel concurrent time i and number of neighbours, compared with an optimal
predictor
channels is effective in reducing channel switching delays. In fact, recent studies on pre-
joining schemes to reduce channel switching delay have considered this variable [150].
In this section I evaluate a scheme where the seven most popular TV channels1 are
pre-joined, and hybrid schemes, where the STB pre-joins both the seven most popular
channels and a subset of neighbours. There are two reasons why I consider here the
seven most popular channels, instead of any other number. First, these are the national,
free-to-air broadcast channels. Second, I tested the scheme considering a different
number of popular channels, and the conclusions to be drawn are the same. The
results can be analysed in Figure 5.8. To make the distinction clearer, I use different
data point types for the hybrid (“popular channels included”) and the neighbours-only
(“popular channels excluded”) schemes. I considered a concurrent channel time of 2
minutes in this analysis.
1Xon“iyzring thz xhvnnzl populvrity vnvly“i“ prz“zntzy in Xhvptzr I =hzxtion ICGCI)C
LI geyuxing xhvnnel xhvnge yelvy
8 neighbours + popular
8 neighbours
6 neighbours + popular
6 neighbours
4 neighbours + popular
4 neighbours
2 neighbours + popular
2 neighbours
Popular channels only
●
●
●
●
15% of all requests
small improvement
0% 25% 50% 75%
Percentage of switching requests
Popular channels ●Included Excluded
Figure 5.8: Percentage of requests that experience no delay by pre-joining the seven
most popular channels, for a channel concurrent time i equal to two minutes. Dif-
ferent data point types are used for the hybrid (“popular channels included”) and the
neighbours-only (“popular channels excluded”) schemes
By pre-joining the seven most popular channels, the number of switching requests
that experience a small channel change delay is less than 15% of all requests. This result
is quite poor when compared with any of the schemes where neighbouring channels are
pre-joined. Also, the hybrid schemes show a very small improvement over pre-joining
the neighbours only. Since pre-joining all these popular channels represents a very
significant increase on the peak access bandwidth (and on the STB processing require-
ments), I conclude that pre-joining the popular channels is not an effective scheme in
reducing channel change delay.
5CK eersonvlisey sxhemeO trvxking user wehvviour
The simple scheme evaluated in Section 5.4 offers interesting results. Notwithstanding
its simplicity, a significant number of requests are to channels available in the STB,
resulting in no zapping delay experienced by the user. But will vll users experience
such benefit? To assess this, I invite the reader to look at Figure 5.9. Here, I show the
results of using the simple scheme presented in Section 5.2, but instead of the average,
as before, I now present the median, 5th and 95th percentile, to understand how the
scheme performs on a per-user basis. It is clear from the figure that the variance is
high. Some users would benefit significantly from using the simple scheme (the “linear
5C6 eersonvlisey sxhemeO trvxking user wehvviour L5
zapping fans”), but others would experience a smaller improvement.
8 neighbours
6 neighbours
4 neighbours
2 neighbours
8 neighbours
6 neighbours
4 neighbours
2 neighbours
8 neighbours
6 neighbours
4 neighbours
2 neighbours
8 neighbours
6 neighbours
4 neighbours
2 neighbours
8 neighbours
6 neighbours
4 neighbours
2 neighbours
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
T
=
1
0
 sec
T
=
3
0
 sec
T
=
6
0
 sec
T
=
1
2
0
 sec
T
=
A
lw
ay
s
0% 20% 40% 60% 80% 100%
Percentage of requests
Figure 5.9: Variance of the percentage of requests that experience no delay by using
the simple scheme for various values of channel concurrent time i and number of
neighbours. In the graph the median, 5th and 95th percentile are presented
My main objective in this section is to try to reduce the variance of Figure 5.9 with-
out decreasing its median. Put in other words, the aim is to improve user experience
for a wider audience, but doing so without affecting the experience of the “linear zap-
ping fans”. For this purpose I devise a personalised scheme. The idea is to track user
actions to build a prediction model. To achieve this each STB will record and maintain
information on both the probability of the user zapping linearly (to maintain the per-
formance for the “linear zapping fans”) plus the probability of him or her switching to
each of the different channels (trying to capture their “favourite” channels, to improve
L6 geyuxing xhvnnel xhvnge yelvy
the performance for other type of users).
This scheme requires two data structures:
1. A channel popularity vector, Ce . This vector maintains a counter for each par-
ticular channel. Every time the user switches to a new channel, its counter is
incremented by one. The favourite channel of a particular user will be the one
corresponding to the maximum value present in the vector.
2. A “jump distance” popularity vector, JDe . This vector maintains a counter for
each particular “jump distance”. Every time a user jumps from channel number
i to channel number j this distance is calculated as j − i and the counter for this
particular distance is incremented by one. A “linear zapping fan” that enjoys
switching mostly to the next and to the previous channel will have maximums at
positions JDe [1] and JDe [−1].
From these two vectors I then get the c TV channels that correspond to the
highest probabilities of the set [Ce ; JDe ]. c is the number of additional channels to
be joined concurrently with the requested one. So, depending on user behaviour, the
STB will either pre-join some neighbours, or some favourites, or a mix of neighbours
and favourites.
Another important aspect of user behaviour is its temporal dynamics. It is known [42,
160, 161] that user behaviour changes with the time of day (for instance, morning v“
evenings), day of the week (weekend v“ weekday), and even period of the year (holiday
v“ working period). Besides this fact, a Set Top Box is usually shared by many people
in one household, and different people may have very different behaviours (a child that
zaps constantly v“ the grandparents that settle in very specific channels). Considering
this, I test the personalised scheme considering two types of vectors:
1. Non-ageing vectors. In this case, the two vectors Ce and JDe do not age.
2. Ageing vectors. In this case, the two vectors Ce and JDe age. The objective is
to capture changes in user behaviour. The vectors age in accordance with Equa-
tion 5.1, an exponential moving average. The coefficient  represents a constant
smoothing factor between 0 and 1. A higher  discounts older observations faster.
d[k] is a vector representing the k -th observation. One of its elements will be
equal to 1 (corresponding to the channel switched to), while the others are all
zero. k [k] represents the value of each counter (in the vector considered) at the
time of the k -th observation.
5C6 eersonvlisey sxhemeO trvxking user wehvviour LL
k [k + 1] = d[k] + (1− )k [k] (5.1)
5CKCF Evvluvtion
As with the previous schemes, this one is evaluated by means of a trace-driven analysis
using the method explained in Section 5.3. I evaluate this scheme for a single value
of the concurrent channel time, 60 seconds. I also restrict the number of pre-joined
channels to 2, 4 and 10. I tested the scheme for various values of , ranging from
0.01 to 0.9. For clarity sake, only a subset of the results is presented in Figure 5.10
(specifically: two additional channels, 60 seconds of concurrent channel time, and six
values for the parameter ).
Aging α = 0.20
Aging α = 0.10
Aging α = 0.07
Aging α = 0.05
Aging α = 0.03
Aging α = 0.01
Non−aging
2 neighbours
●
●
●
●
●
●
●
●
0% 25% 50% 75%
Percentage of requests
Figure 5.10: Variance of the percentage of requests that experience no delay by using the
personalised scheme for a channel concurrent time i equal to 60 seconds, 2 neighbours,
and several values of . In the graph the median and the standard deviation are
presented
As explained before, the main goal of this scheme is to reduce the variance of
Figure 5.9. As a measure of the variability of the results, I analyse the standard
deviation, represented graphically in the figure. As can be observed, the standard
deviation value is basically the same for any scheme tested. The use of this scheme
decreases very slightly the standard deviation, but the decrease is insignificant and is
therefore imperceptible in the figure. In conclusion, this personalised scheme does not
improve user experience when compared with the simple scheme analysed in Section
5.2. This result implies that the benefit of the scheme arises solely from maintaining the
neighbours as an option. There was no gain in adding the favourites option. I speculate
that users that are not “linear zapping fans” are not “zapping fans” in general. It is
LM geyuxing xhvnnel xhvnge yelvy
left as future work to understand if that is the case, and also devising other techniques
to improve the zapping experience for a wider population.
5CL Disxussion
The pre-joining schemes evaluated in this chapter include several positive points:
1. They eliminate channel change delay for a very significant percentage of requests.
2. There is no user perceived picture distortion during the zap process, since no
extra low bandwidth streams are used for the zapping period and no low quality
video is needed.
3. There is no bandwidth increase in the core, regional, or metro networks. Also,
the increase of the average access bandwidth is residual. It is a fact that the peak
bandwidth increases (I discuss this increase below), but since zapping periods are
rare events in the course of a day, the access bandwidth is not affected on average.
4. No changes need to be made to the core of the network, to any network element,
or to the media server. There is no need for extra servers in the network. The
only change needed is an upgrade of the STB software. This is therefore a cheap
solution, very simple to implement.
There is no perfect scheme, of course, and this one is not without its drawbacks:
1. Not all requests are improved, so the user experience will vary: some requests
will experience virtually no delay, while others will experience the normal IPTV
delay, which is high.
2. The access network will need to accommodate the peak bandwidth for sending
several channels in parallel, during zapping periods. In this study I considered
sending between 2 to 8 channels in parallel. As SDTV channels using MPEG-
2 require around 4 Mbps guaranteed bit rate per channel, sending this number
of channels in parallel increases the bandwidth requirements proportionally. The
access networks in most developed countries already offer tens of Mbps of average
download speeds and these are expected to increase in the near future [157], so
this peak bandwidth is attainable on the access link.
3. The Set Top Box needs to be able to synchronise and buffer several TV channels
in parallel, which may increase its cost.
5CM Xonxlusions LN
Overall, the pre-joining schemes studied in this work offer interesting results. They
have, however, the downsides referred to above. To mitigate these drawbacks it is, first
of all, necessary to control the number of TV channels sent in parallel, matching it to
the resources available (this would lessen problems number 2 and number 3 above).
Also, other schemes may be used in parallel with this one to overcome the fact that
user experience is variable (problem number 1). For example, when the user requests
a channel that is not present in the STB, a boost stream could be requested as in
the schemes presented in Chapter 3 (Section 3.2.2). Such hybrid scheme could be an
interesting proposition in terms of quality of experience and cost. User experience
would not vary, since all requests would experience reduced channel change delay. The
cost of the solution would be lower than a pure boost-stream solution as fewer requests
would be sent to the network, due to the fact that a significant percentage of requests
would be served by the STB alone. In consequence, fewer dedicated servers would be
required and the overall solution would be cheaper.
5CM Conxlusions
In this chapter, I investigated a specific type of techniques to reduce channel change de-
lay in IPTV networks, namely, predictive pre-joining of TV channels. In these schemes
each Set Top Box (STB) simultaneously joins additional multicast groups along with
the one that is requested by the user, thus anticipating future user behaviour. To eval-
uate these schemes I have performed an empirical analysis using the dataset described
in Chapter 4. To the best of my knowledge, this is the first empirical study where
channel change delay reduction techniques are evaluated using real IPTV usage data.
In the first scheme evaluated the neighbouring channels (TV channels adjacent to
the requested one) are pre-joined by the Set Top Box during zapping periods, simulta-
neously with the one requested. Thus, in the event of the user switching next to any
of these channels, switching latency is virtually eliminated. As TV users enjoy zapping
linearly — i.e., they tend to switch up and down using the remote control — this scheme
seemed favourable. Indeed, the main conclusion of my analysis is that by using such a
simple scheme, the zapping delay can be virtually eliminated for a very significant per-
centage of channel switching requests. As an example, by sending the previous and the
next channel concurrently with the requested one, for only one minute after a zapping
event, switching delay is eliminated for around half of all channel switching requests.
This is achieved with a negligibly increase of the average bandwidth utilisation in the
access link.
ME geyuxing xhvnnel xhvnge yelvy
I have compared this simple scheme with an ideal predictor, having realised that
its performance was close to the optimal case. However, user behaviour can vary sig-
nificantly, leading to a high variation of the results: although some users would benefit
tremendously from using this simple scheme, others would see only a relatively small
improvement. To address this problem I designed and evaluated other schemes. The
first was to pre-join popular TV channels, but this scheme proved inefficient. The sec-
ond was a personalised scheme where user behaviour is tracked. The results showed
the improvement over the simple scheme was statistically insignificant.
While in this chapter I was concerned with a specific aspect of IPTV user’s quality of
experience, in the next two chapters I will change the focus to the design and operation
of an IPTV network, by proposing novel techniques to increase its resource and energy
efficiency.
Chvpter K
gesourxe vny energy effixient
network
The previous chapter was devoted to the improvement of IPTV users’ quality of expe-
rience. In the next two chapters the focus moves to the design and operation of IPTV
networks. In these chapters I propose novel techniques to increase the resource and
energy efficiency of IPTV infrastructures. The first such technique is based on a simple
paradigm: “Avoid waste!” [167].
IPTV services are bandwidth intensive, requiring low latency and tight control of jit-
ter. To guarantee the quality of experience required by its customers, service providers
opt to build “tvtix multicast trees for the distribution of TV channels. Referring to the
reference architecture presented in Figure 2.5 (Chapter 2), this means vll DSLAMs join
vll multicast groups (they thus receive content of vll TV channels). As particular chan-
nels have no viewers at particular time periods, this method is provably resource and
energy inefficient. In this chapter, I argue that the expected increase in the quantity
and quality of the TV channels distributed in IPTV networks will become a serious is-
sue, bandwidth and energy wise. To alleviate this problem, I propose a yynvmix scheme
where only a selection of TV multicast groups is joined by the network nodes, instead
of all. This scheme is evaluated by means of a trace-driven analysis using the dataset
described in Chapter 4. The objective is to study the tradeoff between the bandwidth
savings of using this technique and the number of requests that will experience higher
channel change delay as a consequence.
I demonstrate that by using the proposed scheme IPTV service providers can save
a considerable amount of bandwidth while affecting only a very small number of TV
channel switching requests. To understand how these bandwidth savings are translated
MG gesourxe vny energy effixient network
into energy savings, I also develop a power consumption model for network equipment
based on real measurements reported recently in the literature. I conclude that while
today the bandwidth savings have reduced impact in energy consumption, with the
introduction of numerous very high definition channels this impact will become signifi-
cant, justifying the use of resource and energy efficient multicast distribution schemes.
KCF Introyuxtion
We have been depleting the natural environment since the times of the industrial revolu-
tion. There is a scientific consensus that our planet will be unable to provide long-term
support if this trend persists. Today, the Internet (excluding home networks, PCs and
data centres) consumes about 0.5% of the current electricity supply of a typical OECD
nation. Although this still represents a relatively small share of the global energy con-
sumption, this fraction is expected to increase quickly [14, 197]. By recognising this
fact, several consortiums are working to improve the ICT sector’s energy efficiency.
GreenTouch [89], for instance, aims to increase network energy efficiency by a factor
of 1000 from current levels by 2015. To accomplish its goal, it focuses on the design of
new network architectures and on the creation of the enabling technologies on which
they are based.
IPTV is a resource intensive service with stringent quality of service requirements.
It requires high bandwidth, low latency and low jitter. As explained in Chapter 2, each
video stream is encoded at a bit rate that can vary from around 4Mbps (SDTV) to 20
Mbps (HDTV). In the future this figure may increase by one or two orders of magni-
tude, with the advent of ultra high definition video standards (2K, 4K, UHDTV) [83].
Besides the increase in the resolution of each TV channel, and its consequent bandwidth
requirements upgrade, the number of TV channels offered is also expected to increase.
AT&T already offers 700 TV channels [160] for their IPTV customers. According to
a recent press release by the European Commission [57], over one thousand channels
have been established in the UK alone until 2009. But recent trends anticipate the
likely growth of the number of TV channels in the near future. Narrowcasting services
— broadcasting to a very small audience [133] —, for example, are growing in impor-
tance. Niche channels are emerging to offer TV services to narrowly targeted audiences
[26, 170]. An extreme example of this type of services was recently announced by Por-
tugal Telecom, the largest telecommunications service provider in Portugal. As part
of its IPTV service, Meo, Portugal Telecom is now offering the possibility for any cus-
tomer to create his or her own TV channel [159]. At the time of writing ten thousand
6CF Introyuxtion MH
TV channels have already been created [60], and some are quite successful. This trend
is expected to continue in the future as there seems to be clear market opportunities
in offering this type of long tail service [12]. This calls for novel, efficient distribution
schemes.
Unfortunately, current IPTV networks are not efficient. Service providers opt to
distribute all TV channels, continuously, everywhere. Referring to the network archi-
tecture presented in Figure 2.5 (Chapter 2), this means vll DSLAMs join vll multicast
groups. Originally, IP multicast had a dynamic nature, but IPTV providers opted for
“tvtix multicast to guarantee the quality of experience required by its users (i.e., to
guarantee a small channel change delay, as explained before), and to reduce service
complexity (in terms of state and control traffic overheads). But static multicast is
provably inefficient, as is demonstrated below. As soon as the number of channels sur-
passes the number of users at a certain access node, sending all channels is wasteful.
Also, as already mentioned in Chapter 4 (Section 4.2.4), recent work [42, 161] has shown
that channel popularity is highly skewed (following a Zipf-like distribution). While a
small number of channels is very popular, dozens or even hundreds of TV channels are
very rarely watched. Service providers seem to recognise this problem and are already
concerned with the efficiency of their IPTV networks [40].
By analysing the dataset described in Chapter 4, I indeed observe that for the
most popular channels there is always at least one viewer per access node (DSLAM),
at any one time. In this chapter I call a channel that has at least one viewer in a
particular network node (be it a DSLAM or a router) an vxtive xhvnnel in that
particular node. Despite some channels always having viewers, the number of active
channels in each DSLAM is rarely above 60. This can be seen in Figure 6.1. This
graph presents the average number of actives channels in every DSLAM and regional-
core router in the network, as a function of the number of users. This figure is based on
analysis of the whole dataset (255 thousand users, 6 months). It can be clearly observed
that the network is wasting resources by distributing all 150 TV channels everywhere.
In the figure, the shaded area represents the bandwidth savings opportunities for an
IPTV service provider. With the likely increase of the number of channels (and of its
bandwidth requirements) this inefficiency may become problematic.
Considering the above, I propose to reduce these inefficiencies by not building static
multicast trees, i.e., not distributing all TV channels, continuously, everywhere. Instead
of the network nodes joining all multicast groups, I propose each node to join only a
limitzy “zlzxtion of channels. A relevant point about this scheme is its dynamic nature.
This is important due to channel popularity dynamics [161]. This design goal is fulfilled
MI gesourxe vny energy effixient network
0 
20 
40 
60 
80 
100 
120 
140 
160 
10 100 1000 10000 100000 
N
u
m
b
e
r 
o
f 
ch
a
n
n
e
ls
 
Number of users 
Channels with viewers 
Channels broadcasted 
DSLAM 
regional-core 
router 
Figure 6.1: Average number of active channels (TV channels with viewers) per network
node (including DSLAMs and core-regional routers). Nodes are ordered by the number
of users they serve
by each network node joining only the active TV channels plus a small subset of the
inactive ones. This scheme is dynamic because the list of joined channels varies with
user activity.
To evaluate the proposed scheme I perform a trace-driven analysis on the dataset
described in Chapter 4. I evaluate this scheme in two ways. First, I analyse the tradeoff
between the bandwidth savings of using this technique and the number of requests
affected. A request is considered affected when the user requests a channel that in that
particular moment is not part of the “selected channels” list of that node (i.e., that node
has not joined that particular multicast tree). In such case, the user will experience a
higher-than-usual channel change delay. This is due to the fact that a join message to
that multicast channel will have to go up towards the source until it finds the nearest
leave of the multicast tree. As I explained in the previous chapter, channel change delay
is a problem in IPTV networks, so it is important the number of affected requests to
be as low as possible to avoid jeopardising service quality. In addition, reintroducing
dynamics in the multicast network reintroduces protocol overhead. Control messages
start flowing in the network as multicast trees are joined and pruned. This also calls
for a tight control over the number of requests affected. Second, I analyse how these
bandwidth savings translate into energy savings. For this purpose I create a power
consumption model for routers, based on real measurements [184], and evaluate it
considering several scenarios.
The results show that it is possible to significantly reduce the bandwidth used by
6CG helextive joining M5
IPTV services in the network, while affecting only a very small number of switch-
ing requests. Considering current IPTV service offerings (hundreds of SD or HD TV
channels), these savings have negligible impact on energy consumption. Considering
futuristic scenarios, with IPTV offers of thousands of very high definition TV channels,
the savings of using such dynamic multicast scheme become meaningful. The relative
advantage of the proposed scheme is even more significant if one considers the use of
equipment with more energy-proportional [24] power profiles.
The rest of the chapter is organised as follows. In Section 6.2 I present the scheme
proposed in this chapter: selective joining. I then present the methodology used to
evaluate this scheme in Section 6.3, and evaluate it in Section 6.4. I analyse the impact
of using the proposed scheme on energy consumption in Section 6.5. The effects of
using this scheme on channel change delay are briefly discussed in Section 6.6, and the
chapter closes with Section 6.7.
KCG helextive joining
Currently, IPTV operators distribute all TV channels continuously everywhere, in order
to minimise channel change delay. All access nodes in the network join all TV multicast
groups. This means that all DSLAMs are leafs of the multicast tree of every TV channel.
My proposal, “zlzxtivz joining, is for each node to join only v “uw“zt of the complete
selection of TV channels. Namely, each node will join:
1. the channels for which there are viewers (the active channels) plus
2. a small subset of inactive channels. Inactive channels have no viewers in the node
under consideration. I call the number of inactive channels that are joined by the
node the “izz of thz invxtivz joinzy “zt, or fkactfse\pet\pfwe.
This scheme requires a single data structure to be maintained at each network node,
containing two elements: one to store information on the joined channels, jlfkea\pet,
and another to record the number of viewers for each joined channel, krj\sfeteop.
An inactive channel will have its corresponding krj\sfeteop variable equal to zero.
The proposed scheme is a form of hybrid multicast. The interaction between the
user and the DSLAM is still via normal IGMP, but static multicast is replaced by a semi-
dynamic multicast in the IP network. The dynamic nature comes from the jlfkea\pet
always including the active channels. For this reason, the “selected channel” list changes
yynvmixvlly with user demand. At the same time, the channels that have an higher
M6 gesourxe vny energy effixient network
probability of being watched in the future are “automatically” joined. Popular channels
are the ones people watch more, hence there is a high probability of them having at least
one viewer, which means they are usually included in the jlfkea\pet. In summary,
a largely static group of popular channels will be kept in the list while a dynamic
group of less popular channels leaves and joins with channel popularity dynamics.
Selective joining can thus be seen as a form of cross-layer optimisation, using user level
information about content popularity to drive the hybrid protocol.
KCH bethoyology
The scheme proposed in this chapter is evaluated by means of a trace-driven analysis.
The IPTV trace detailed in Chapter 4 is used as input to the analysis performed. I
analyse the use of the proposed scheme in two particular nodes in the network topology
(I refer the reader to Figure 2.5 again): the DSLAMs and the core-regional routers.
The reason why I chose these two particular locations is related to the information I
can extract from the data trace, as will be made clear in the following.
Recall from Chapter 4 (Section 4.2.3) that I parsed the raw IPTV trace data to
create a single file pzr hiW. Each file includes all channel switching requests made by
the users of a specific household. The first line of each file included the DSLAM IP
address users send their IGMP signals to. With this information I am able to create
a single time-ordered trace file that includes all switching events sent to a “pzxifix
DhaVb. This allows the evaluation of this scheme at the DSLAM level. An example
of the new version of the trace, for a particular DSLAM, follows.
Jri . --71.703yRMy/0y.-.41.26.65
Jri . --71.711yALTNy/0y.-.41.26.65
Jri . --71.711yRMy/2y.-.41.44..-.
Jri . --71.715yRMy.5/y.-.41.5-.5-
Jri . --71.715yRMy/0y.-.41../-..
Jri . --71.72.yALTNy.5/y.-.41.5-.5-
Jri . --71370/yALTNy/0y.-.41../-..
Jri . --7147-1yALTNy/2y.-.41.44..-.
As can be seen, the difference from the parsed version illustrated in Chapter 4
(Section 4.2.3) is the inclusion of information on the STB that sent the IGMP request.
This information allows me to understand how many viewers each channel has at each
moment, pzr DhaVb.
6CH bethoyology ML
I now refer the reader again to the reference network topology presented in Figure
2.5. The IP network usually has a two-level, hierarchical structure [79]: the regional and
the core networks. That is the case of the network where the IPTV traces were collected
from. One core-regional router1 aggregates all traffic from a specific region. Thus, the
traffic sent from all DSLAMs destined to the core goes through this router. Unfor-
tunately, the IPTV traces do not include information about the region each DSLAM
belongs to. Fortunately, there is very easy way to deduce this information from the
data. The IPTV service studied includes as part of its channel bundle one or two re-
gional channels pzr rzgion. In the trace, each of these channels is numbered differently
from the others. Instead of being just a single number, it also includes information
on the region it is distributed on. For example, channel numbers 5-J and 6-J are the
regional channels from Madrid. With this information I am able to create a single time-
ordered trace file that includes all switching events sent to all DSLAMs in a “pzxifix
rzgion. This allows the evaluation of this scheme at the core-regional router level.
To evaluate the proposed scheme I developed a Python script that checks each line
of the input trace, to obtain each switching event received by each node (a DSLAM
or a core-regional router). The current switching event is analysed, and one of these
actions is performed:
1. If it is an RM event, and
a if the channel is in the jlfkea\pet, the channel was joined by the node before.
The global counter eft and this channel’s counter krj\sfeteop are incre-
mented. The counter eft counts the number of requests served quickly by this
node.
b if the channel is not in the jlfkea\pet, the channel was not pre-joined by the
DSLAM. The global counter jfpp is incremented. The counter jfpp counts the
number of requests that are not served quickly by this node. The node has to
send the join message towards the source which increases the channel change
delay. The channel number is added to the jlfkea\pet, and its krj\sfeteop
is set to 1.
2. If it is a ALTN event, decrement this channel’s krj\sfeteop counter. Then,
a if the number of viewers is above zero, keep the channel in the jlfkea\pet.
1In fvxt thi“ routzr i“ rzplixvtzy for rzlivwility vny yzpznyvwility rzv“on“C
MM gesourxe vny energy effixient network
b if the number of viewers is equal to zero, the channel is inactive. In that
case, if the number of inactive channels is below the fkactfse\pet\pfwe, keep
the channel in the jlfkea\pet. Else, choose one of the inactive channels and
remove it from the jlfkea\pet. In this case, the DSLAM sends an IGMP leave
message from this channel towards the source.
core IP network
IPTV
head-end
core
network
regional
network
metro
network
x
Jul 1 00:41:36|UP|23|10.74.59.98
Jul 1 00:41:44|UP|25|10.74.77.101
Jul 1 00:41:48|UP|23|10.74.120.1
Jul 1 00:42:51|DOWN|23|10.74.59.98
Jul 1 00:43:48|UP|182|10.74.80.80
Jul 1 00:44:51|DOWN|182|10.74.80.80
Jul 1 00:46:32|DOWN|23|10.74.120.1
Jul 1 00:47:04|DOWN|25|10.74.77.101
Parsed log file
1.
2.
3.
4.
5.
6.
7.
8.
input
x1
x2
x1_1
x1_2
x3_1
x3_2
x = {all TV channels}
xi = subset of x
xi_j = subset of x1
x3
x3_3
x3_4
x1_3 x1_4
x3_4.joined_set = {23}
x3_4.num_viewers = {1}
x3_4.joined_set = {23,25}
x3_4.num_viewers = {1,1}
x3_4.joined_set = {23,25}
x3_4.num_viewers = {2,1}
x3_4.joined_set = {23,25}
x3_4.num_viewers = {1,1}
x3_4.joined_set = {23,25,182}
x3_4.num_viewers = {1,1,1}
x3_4.joined_set = {23,25,182}
x3_4.num_viewers = {1,1,0}
x3_4.joined_set = {23,25}
x3_4.num_viewers = {0,1}
x3_4.joined_set = {23}
x3_4.num_viewers = {0}
Contents of structure x3_4 at each step
1.
2.
3.
4.
5.
6.
7.
8.
Figure 6.2: Proposed methodology
Figure 6.2 illustrates the proposed methodology with a simple example1. I assume
the sample trace is from a DSLAM, and that the fkactfse\pet\pfwe is equal to 1. This
means there can be only one TV channel without viewers in this DSLAM. At 12:41:36am
the DSLAM receives a join message for channel 23 from the STB with IP address
.-.41.26.65. As the channel is not in the jlfkea\pet (maintained in the structure
1ihi“ figurz i“ wv“zy on thz rzfzrznxz vrxhitzxturz prz“zntzy in [igurz GCJC
6CI Zvvluvtion MN
x3 4 in the figure), the global counter jfpp is incremented, and the DSLAM sends the
join message towards the source. The channel number is added to the jlfkea\pet x3 4,
and its krj\sfeteop counter is set to 1. This can be observed from structure x3 4’s
contents, step 1. Then, at 12:41:44am, the STB with IP address .-.41.44..-. sends
a join message to channel 25, and a similar procedure occurs. Four seconds later the
STB with IP address .-.41../-.. sends a join message to channel 23. As the channel
is in the jlfkea\pet, this time it is the global counter eft that is incremented. This
channel’s krj\sfeteop counter is incremented (to 2). When the down message to this
channel arrives at the DSLAM from STB .-.41.26.65, its counter is decremented to 1.
As there are still viewers the channel is kept in the jlfkea\pet. When the new IGMP
down message to this channel arrives at the DSLAM, at 12:46:32am (step 7), its counter
krj\sfeteop is decremented to zero, which means there are no viewers for this channel.
It became inactive. At that time, there is another inactive channel in the jlfkea\pet
(channel 182; step 6). As there are two inactive channels and the fkactfse\pet\pfwe
is equal to 1, one of these two channels is removed. In this example I assume the least
recently watched channel is removed from the jlfkea\pet list (channel 182).
KCI Evvluvtion
As I said in the previous section, to evaluate the proposed scheme I perform a trace-
driven analysis on the dataset presented in Chapter 4. All results I present in this
section arise from the analysis of the whole data set (6 months, 255 thousand users).
I test two schemes to decide which channel to remove from the jlfkea\pet when a
channel becomes inactive. The first is to remove an inactive channel randomly. The
second is to remove the least recently watched channel. The two schemes produce
indistinguishable results, so for clarity sake only the results from one scheme — the
random — are shown.
In Figure 6.3 I present a graph with the number of TV channels joined by each
node (fl -axis) as a function of the number of inactive channels that are joined by the
node (the fkactfse\pet\pfwe, i in the figure). The figure presents the median, 10th-
and 90th-percentile. I consider two types of nodes (for the reasons explained before):
DSLAMs and core-regional routers. At the DSLAM level I present the results for five
different values of the fkactfse\pet\pfwe, i. As can be seen, not joining all 150 TV
channels represents bandwidth savings in the network. With an fkactfse\pet\pfwe of
only 20 TV channels, bandwidth can be reduced by 50%. At the regional level (“core-
reg” in the figure) I present the results considering i = 0 only, i.e., the core-regional
NE gesourxe vny energy effixient network
All channels
i = 0 (core−reg)
i = 25 (DSLAM)
i = 20 (DSLAM)
i = 15 (DSLAM)
i = 10 (DSLAM)
i = 5 (DSLAM)
i = 0 (DSLAM)
●
●
●
●
●
●
●
●
33% bandwidth
reduction
50% bandwidth
reduction
0 20 40 60 80 100 120 140 160
Number of channels joined
Figure 6.3: Number of channels joined when using the selective joining scheme for
various values of the inactive set size
routers join only the active channels1. In this case, the average bandwidth savings are
equal to 33%. In the figure, I also compare the proposed scheme with the one currently
used by IPTV providers (“All channels”).
In order to analyse the effect these schemes will have on the quality of experience
of IPTV users, I now inspect the percentage of requests to channels not joined by the
node. As explained before, in this case the node has to send a join message to this
channel towards the source which increases the channel change delay. This percentage is
calculated as the value of the counter jfpp divided by the total number of requests. The
results are shown in Figure 6.4, again for various values of i. For an fkactfse\pet\pfwe
of 20, the percentage of requests affected at the DSLAM level is less than 2% on
average. At the core-regional level this figure is almost negligible. In core-regional
routers, joining active channels only thus seems a good option.
A decrease in the number of TV channels joined by a node represents bandwidth
savings. By joining fewer channels, the nodes processes, and the links transport, less
bits. How significant these savings are, both in resource and energy terms, is therefore
intrinsically dependent on the data rate at which channels are distributed. To clearly
understand the significance of the savings achieved by the proposed scheme, I analyse
three scenarios, characterised in Table 6.1.
In the first scenario, 150 TV channels are distributed in Standard Definition format
(SDTV). This represents the IPTV service offering under analysis, at the time of trace
1ihz rzv“on fihy I prz“znt only thi“ vvluz fiill wzxomz xlzvr in thz nzxt pvrvgrvphC
6C5 Impvxt on energy xonsumption NF
All channels
i = 0 (core−reg)
i = 25 (DSLAM)
i = 20 (DSLAM)
i = 15 (DSLAM)
i = 10 (DSLAM)
i = 5 (DSLAM)
i = 0 (DSLAM)
●
●
●
●
●
●
●
●
<0.1% requests
affected
<2% requests
affected
0% 2% 4% 6% 8% 10% 12% 14% 16%
Percentage of requests
Figure 6.4: Percentage of requests affected for various values of the inactive set size.
Table 6.1: Description of the three scenarios
Scenario Media format Bit rate TV channels Wvnywiyth svvings
150SD SDTV 4 Mbps 150 ECH Gwps
700HD HDTV 20 Mbps 700 L Gwps
3kUHD 4K 200 Mbps 3000 HEE Gwps
collection. A bandwidth saving of 50% means reducing load by 300 Mbps. This is not
very significant. Currently, however, most IPTV networks already offer more channels
(AT&T offers 700 [160]) in high definition (HDTV). Assuming such scenario the band-
width decrease now accounts to around 7 Gbps. Looking further into the future, as
explained in section 6.1, one can anticipate many more channels and even higher quality
streams (digital cinema standard 4K, for instance, or UHDTV [83]). In the futuristic
scenario I therefore assume 3000 4K TV channels. In this case, the bandwidth savings
are already very significant, with a magnitude of several hundred Gbps.
KC5 Impvxt on energy xonsumption
Saving bandwidth may be, pzr “z, an important objective. Nevertheless, in this chapter
I additionally analyse the impact the proposed scheme has on energy consumption. In
principle, bandwidth savings should result in energy savings. Less bits need to be
transported in the links, and less bits need to be processed by the routers. Also,
reducing load in the network offers more opportunities to put some equipment to sleep
or to adapt line rates, in order to save energy. In this section I try to understand if the
NG gesourxe vny energy effixient network
bandwidth savings reported in the previous section are translated into relevant energy
savings.
KC5CF eower xonsumption moyel
To be able to quantify the energy savings achieved by using the selective joining scheme,
in this section I build a power consumption model of a network node. Several factors
affect the power consumption of such node [136]:
1. Base chassis power. This is the power to maintain the chassis on. It is a fix
amount independent of load, including the power consumed by components such
as fans, memory, etc.
2. Number of active linecards. A linecard is the electronic circuit that interfaces
with the network.
3. Number of active ports in each linecard.
4. Port capacity. This is the line rate forwarding capacity of individual ports.
5. Port utilisation. This is the actual throughput flowing through a port, relative
to its capacity.
Based on these variables, I use the following model of power consumption e of a
router.
e = ech +
L∑
i20
eli (6.1)
In equation 6.1 ech refers to the power consumption of the chassis. a is the number
of linecards that are active, and eli is the power consumption of linecard i. The
power consumption of each linecard is calculated based on the model proposed by
Sivaram zl vlC [184] for a NetFPGA card, and is presented as Equation 6.2. By using a
high-precision hardware-based traffic generator and analyser, and a high-fidelity digital
oscilloscope, the authors devised a series of experiments allowing them to quantify the
per-packet processing energy and per-byte energy consumption of such linecard.
el = ec +KeE +cIEp +gEb (6.2)
In this equation:
6C5 Impvxt on energy xonsumption NH
Table 6.2: Linecard power profile
Energy component and description Estimate from [184].
Power consumed by unconnected linecard card (ec) 6.936 W
Power consumed per connected Ethernet port (eE) 1.102 W
Per-packet processing energy (Ep) 197.2 nJ
Per-byte energy (Eb) 3.4nJ
• ec is the constant baseline power consumption of the NetFPGA card (without
any Ethernet ports connected).
• K is the number of Ethernet ports connected.
• eE is the power consumed by each Ethernet port (without any traffic flowing).
• cI is the input rate in packets per second (pps).
• Ep is the energy required to process each packet.
• g is the traffic rate in bytes per second. I am assuming the input rate is equal to
the output rate.
• Eb is the total per-byte energy. This includes the energy required to receive,
process and store a byte on the ingress Ethernet interface; and the energy required
to store, process and transmit a byte on the egress Ethernet interface.
The inputs to this model are presented in table 6.2, again based on the measure-
ments reported in [184].
In Figure 6.5 I present the power consumption of a router based on this model,
and assuming ech = 430l . This value for the power consumption of the chassis is
based on power measurements of the Cisco GSR 12008 router, performed by Chabarek
zt vlC [44]. This is the power profile for a router with four linecards with 4x1Gbps ports
each. The plot presents power consumption as a function of traffic load. Note that the
y-axis present values from y = 400[l ] to y = 500[l ] only. This is therefore a zoomed
version of the power profile. The reason why I present it this way first is the fact
that current network equipment is not energy proportional [24]. The baseline power
(from maintaining the chassis powered on) is very high and is, by a large margin,
the main component of router power consumption. But this zoomed version allows
the observation of these relatively significant power consumption “jumps” at regular
intervals. The small jumps represents turning on a new Ethernet port in the linecard,
while the bigger jumps represent turning on one linecard.
NI gesourxe vny energy effixient network
400
420
440
460
480
500
Turn on linecard
Turn on Ethernet port
0 4000 8000 12000 16000
Traffic load (Mbps)
T
o
ta
l 
p
o
w
er
 c
o
n
su
m
p
ti
o
n
 (
W
)
Figure 6.5: Power consumption model (zoomed)
To contextualise, Figure 6.6 shows the previous figure zoomed out. This figure
clearly illustrates how far away current routers are from an energy-proportional be-
haviour. Anyway, with the ongoing green research novel network devices having lower
energy when idle are expected in the future. I consider this trend in the analysis that
follow.
KC5CG gesults
I now analyse how the bandwidth savings reported in Table 6.1 translate into energy
savings. As shown in Section 6.4, the scheme proposed in this chapter — selective
joining — allows an IPTV provider to reduce its network bandwidth consumption
without affecting user experience significantly. In the analysis I consider the use of
the selective joining scheme in an IPTV network with the following configuration.
The DSLAMs join the active channels plus 20 inactive channels (i.e., they set their
fkactfse\pet\pfwe to 20), and the routers join only the active channels (i.e., they set
their fkactfse\pet\pfwe to 0). As illustrated in Figure 6.3, this represents an average
traffic decrease of 50% and 33% to the DSLAMs and core-regional routers, respectively
(while maintaining an acceptable quality of experience, as can be attested in Figure
6.4). Assuming such scenario, the traffic decrease in the regional network (I again refer
the reader to the reference architecture, Figure 2.5) would be between 50% (decrease in
DSLAMs load) and 33% (decrease in core-regional router load). In the core network,
6C5 Impvxt on energy xonsumption N5
0
100
200
300
400
500
Model
Energy proportional node
0 4000 8000 12000 16000
Traffic load (Mbps)
T
o
ta
l 
p
o
w
er
 c
o
n
su
m
p
ti
o
n
 (
W
)
Figure 6.6: Current router power consumption v“ energy-proportional node
the traffic decrease would vary between 33% (core-regional router) and 20%. The jus-
tification for these 20% is given in Chapter 7. In that chapter I demonstrate that at
any particular moment an average of one fifth of the TV channels does not need to be
distributed in the IPTV network, as they do not have a single viewer.
I consider the three scenarios presented in Table 6.1 in the analysis: .2-PA, 4--EA
and 0hREA. For the first scenario, I assume a router with four linecards with 4x1Gbps
Ethernet ports each, as in Figure 6.6. For the second scenario I scale up the node to
sixteen linecards of the same type, for it to be able to handle the increased aggregate
throughput. The capacity of each node is now assumed to be equal to 64Gbps. The
capacity of the nodes of the third scenario has to scale up to the Tbps range. I assume
fourteen 4x40Gbps linecards for an aggregate capacity of 2.2Tbps. This is a different
type of linecard from the one measured by Sivaram zt vlC [184]. I therefore assume a
4x40Gbps linecard presents the same power profile as forty 4x1Gbps.
A final note concerning the assumptions made in this analysis. For their infrastruc-
tures to be reliable and to provide high performance, network providers build densely
interconnected networks with many redundant paths [44]. In their networks, pairs of
routers are typically connected by multiple physical cables that form one logical bun-
dled link [69]. I therefore assume that any pair of routers will maintain multi-bonded
channels to inter-communicate. I also assume the routers will use each of these paral-
lel channels to its full capacity before deciding to use a new free channel, i.e., before
N6 gesourxe vny energy effixient network
turning on a new port/linecard.
0%
10%
20%
30%
40%
50%
linecard turned on
in baseline scenario linecard turned on in selective joining scenario
25% 35% 45% 55% 65% 75%
Baseline traffic load (%)
P
o
w
er
 s
av
in
g
s 
(%
)
3kUHD_ep (max sav) 3kUHD_ep (min sav)
3kUHD (max sav) 3kUHD (min sav)
700HD (max sav) 700HD (min sav)
150SD (max sav) 150SD (min sav)
Figure 6.7: Power savings after introducing the proposed scheme as a function of the
baseline traffic load in the node (regional network)
The power savings for the three scenarios under consideration are presented in
Figures 6.7 and 6.8, for the regional and core network, respectively. The graphics
illustrate the relative power savings of using the selective joining scheme as a factor of
the baseline traffic load according to equation 6.3. The baseline traffic load is the load
of a node that does not use the proposed scheme. This load obviously includes IPTV
traffic.
e (wvseline)− e (selextive joining)
e (wvseline)
∗ 100 (6.3)
In Equation 6.3, e (wvseline) is power consumption at baseline traffic load, whereas
e (selextey joining) is power consumption when using the proposed scheme (a lower
value due to the decrease in IPTV traffic). In the figures I present results for baseline
load values varying from 25% to 75%.
The results for the regional and core networks are presented separately, but as
the conclusions to be drawn are basically the same, their analysis is made jointly.
For each scenario I present a line with maximum and minimum power savings, for
the reasons explained in the paragraph that opened this subsection. The dashed lines
represent the minimum savings obtained in each scenario, while the solid lines represent
maximum savings. Each scenario is represented in different colours. The first conclusion
one can draw from the observation of the graphs is that for current scenarios the
6C5 Impvxt on energy xonsumption NL
0%
10%
20%
30%
40%
50%
25% 35% 45% 55% 65% 75%
Baseline traffic load (%)
P
o
w
er
 s
av
in
g
s 
(%
)
3kUHD_ep (max sav) 3kUHD_ep (min sav)
3kUHD (max sav) 3kUHD (min sav)
700HD (max sav) 700HD (min sav)
150SD (max sav) 150SD (min sav)
Figure 6.8: Power savings after introducing the proposed scheme as a function of the
baseline traffic load in the node (core network)
bandwidth savings achieved with the proposed scheme will have negligible impact on
energy consumption. In the .2-PA scenario, the power savings would represent less
than 1% on average, while in the 4--EA scenario they would increase to just around
3%.
The most interesting results occur in a scenario with many TV channels at high
resolutions, as the 0hREA one. In particular under normal traffic load conditions, with
values below 30%1. The advantage of using the proposed scheme seems clear. Consid-
ering such load conditions, the power savings are on the order of 30% in the regional
network, and 20% in the core. The decrease observed as load increases was expected. As
traffic load increases, the baseline power consumption (the divisor) increases faster than
the relative power reduction (the dividend), and therefore the relative power reduction
gain (the quotient) decreases. This is true in all scenarios, but is less pronounced in
the first two, .2-PA and 4--EA, because the resulting power savings are small. For this
reason this trend is only perceptible in the futuristic scenario.
One aspect that deserves explanation is the lines in the plots not being completely
smooth (the little “steps”). This is particularly evident in the first two scenarios,
.2-PA and 4--EA. The reason is that the fl -axis represents the baseline traffic load in
the node (without using the proposed scheme), while the power savings arise from the
1cztfiork“ vrz provi“ionzy vt prz“znt for thz fior“t xv“z vny mvny timz“ ovzrprovi“ionzyC ihi“ i“
v yz“ign xhoixz from nztfiork opzrvtor“ thvt vllofi“ thzm to protzxt thzir nztfiork“ vgvin“t multiplz
fvilurz“A to hvnylz trvffix vvrivwility vny to “upport thz rvpiy grofith of trvffix volumzC Vxxorying to
v mzv“urzmznt “tuyy of hprint’“ wvxkwonz nztfiork prz“zntzy “omz yzvr“ vgo pFFFrA v vzry “ignifixvnt
pzrxzntvgz of thz nztfiork link“ =KN:) lctcp zxpzriznxzy v lovy vwovz HE: in thz pzrioy vnvly“zyC
NM gesourxe vny energy effixient network
new traffic load (using the proposed scheme) being lower. The power saving peaks that
appear in the graph represent transition points, when a particular event that increases
significantly the energy consumption occurs: in this case, when an additional linecard
needs to be turned on. For instance, in the .2-PA scenario there is a peak precisely in
the middle of the plot. This is because a 50% load in that scenario represents a data
rate equal to 4Gbps. At this point, the network node has to turn on a new linecard
(recall that I am assuming 4x1Gbps linecards). With the proposed scheme, the network
load would be lower than the baseline traffic load, a bit under 4Gbps. So the linecard
does not need to be turned on yet. While the traffic load does not increase over that
transition point the proposed scheme therefore presents a higher-than-average power
saving advantage. In the 4--EA scenario the same occurs, but more frequently. This is
due to the fact that in this scenario the network nodes have eight times more linecards,
so the effect occurs eight times more than in the .2-PA case. A similar effect occurs in
the futuristic scenario. But, as the baseline power consumption is much higher than in
the first two, the bumps are less pronounced, and are hence imperceptible in the figure.
In the plots I also include, for the futuristic scenario 0hREA, the situation where
all routers are energy-proportional (EP) (0hREA\em). These nodes have the energy-
proportional traffic profile presented in Figure 6.6. In this case, the energy saving
advantage from using the proposed scheme is even more pronounced. Note that by
looking at the EP model from Figure 6.6 one would expect the power savings to be
huge when compared with current routers’ power profiles. This is a fact, but at a first
glance this does not look to be the case in the graphs just presented. The reason for
this is the type of analysis I am making here. To make it absolutely clear, the blue lines
show the improvement of using the proposed scheme considering the power profile of
current routers over not using it (assuming that same profile of course). The green lines
show the improvement of using this scheme assuming EP nodes over not using it (again,
v““uming Ee noyz“). Having EP nodes will vlfivy“ decrease the power significantly,
even if the proposed scheme is not being used. That is a fact that can be easily inferred
from Figure 6.6. But the important aspect to emphasise is that using the selective
joining scheme leads to a higher relative gain considering that different starting point
in the analysis (i.e., the use of EP routers).
KCK DisxussionO effext on xhvnnel xhvnge yelvy
The scheme proposed in this chapter represents a tradeoff between channel change delay
and operational efficiency. As explained, if the user requests a channel that was not
6C6 DisxussionO effext on xhvnnel xhvnge yelvy NN
joined previously by a network node, the request has to go up towards the source to
the nearest branch of the multicast tree. It will therefore experience a larger than usual
delay, due to an increase in the network delay. This problem is mitigated by two factors.
First, and as explained in Section 2.1.1, the network delay is a small contributor to the
overall zapping delay in IPTV: buffering and stream synchronisation are the largest
contributors to this delay. Second, if one of the main objectives of the proposed scheme
is satisfied — namely, to affect a very small number of channel requests — the number
of signalling messages transported in the network will not be significant, and hence
overall network delay may not be seriously affected. And, as shown in this chapter,
in particular in Figures 6.3 and 6.4, a significant increase in resource and associated
energy efficiency is possible while affecting just a very small number of channel change
requests.
Anyway, a small contradiction becomes clear. The conclusion of the previous chap-
ter is that zapping delay may be removed through the addition of more channels,
whereas the conclusion of the current chapter is that the removal of unused channels
from the aggregate bundle provides an avenue for saving power. Clearly there is a
trading relationship between the zapping delay and the need to save power. To make
this relationship explicit I present a simple quantitative analyses in this section.
I consider the four scenarios depicted in figure 6.9. Scenario 0 represents the way
IPTV networks operate today. As explained in Section 2.3, static IP multicast is
used, with each DSLAM joining all TV multicast groups and thus receiving content
from all TV channels. The DSLAM then distributes a single TV channel to each Set
Top Box (STB). Scenario 1 is the proposal presented in Chapter 5: pre-joining some
neighbouring channels to reduce channel switching delay. For the analysis I assume
that two channels are pre-joined and that the STB does not leave these groups while
the requested channel is being watched (according to the notation used in the previous
chapter, ceighwours = 2 and concurrent channel time i = vlwvys). In this case, 55%
of all switching requests will experience no delay (Figure 5.4). Scenario 2 is the proposal
presented in the current chapter: selective joining to increase operational efficiency. The
DSLAM does not join all TV multicast groups, but only a selection of the available
channels. In this case, I assume that only the active channels are joined. Recall that
fewer channels joined by the DSLAM represents a reduction in bandwidth and energy
consumption, as analysed in this chapter. Also note that in the current chapter I have
assumed thus far that only one TV channel is distributed to each STB at any one
time (i.e., so far the pre-joining scheme from the previous chapter was not used jointly
with the scheme presented in this chapter). Finally, in scenario 3 I consider the two
FEE gesourxe vny energy effixient network
core IP networkIPTV network w/o
selective joining
STB w/o
prejoining
neighbours
[ch]
(a) Scenario 0
core IP networkIPTV network w/o
selective joining
STB with
prejoining
neighbours
[ch-1, ch, ch+1]
(b) Scenario 1
core IP networkIPTV network with
selective joining
STB w/o
prejoining
neighbours
[ch]
(c) Scenario 2
core IP network
IPTV network with
selective joining
STB with
prejoining
neighbours
[ch-1, ch, ch+1]
(d) Scenario 3
Figure 6.9: The four scenarios considered
proposals from the previous and the current chapter together. I assume the DSLAM is
using the selective joining technique and the STB is pre-joining two neighbours along
with the requested channel.
To quantify the trading relationship between the switching requests that experience
no delay and the number of channels distributed to a DSLAM, I wrote a simple simula-
tion experiment in C. The objective of the simulation is to quantify how many channels
are active, in the DSLAM, on average, at any one time. Note that this simulation is
only performed for scenarios 2 and 3. In the other two scenarios all TV channels are
always distributed to the DSLAM so no simulation is necessary. The input to the simu-
lation was the long term distribution of channel popularity I obtained empirically from
the analysis of the dataset (Figure 4.4). I generate a random number based on this
distribution for each STB in order to simulate what TV channel the user is watching,
and therefore what channels the DSLAM is distributing to the STB. In scenario 2 only
6C6 DisxussionO effext on xhvnnel xhvnge yelvy FEF
the channel the user is watching is distributed, while in scenario 3 several channels
are distributed to the STB: the one the user is watching, the previous and the next1.
This is performed for every STB covered by a single DSLAM. This way it is possible
to quantify how many channels are active, on average, in a single DSLAM. For these
scenarios I consider that the DSLAM serves 409 households (i.e., STBs), which was the
average number of STBs a DSLAM served in Telefonica’s IPTV network at the time
of data collection. I also consider that the network distributes 150 TV channels, which
is this provider’s service offering. I run the simulation 1000 times, and calculate the
median, 90th and 10th percentile. The results are presented in Figure 6.10.
current IPTV
networks
scenario 1
(chapter 4)
scenario 2
(chapter 5)
scenario 3
(ch. 4 + ch. 5)
effect of reducing
concurrent channel time
lower
is
better
higher is better
0
50
100
150
0% 25% 50% 75% 100%
Percentage of switching requests that experience no delay
N
u
m
b
er
 o
f 
ch
an
n
el
s 
jo
in
ed
 b
y
 t
h
e 
D
S
L
A
M
Figure 6.10: Trading relationship between the switching requests that experience no
delay and the number of channels distributed to a DSLAM.
In scenario 0 all 150 TV channels are distributed to the DSLAM, and all switch-
ing requests experience the normal IPTV delay. Therefore, the number of switching
requests that experience no delay is zero in the plot. In scenario 1, again, all TV chan-
nels are distributed to the DSLAM. However, as predictive pre-joining is used, 55% of
all switching requests experience no delay (according to figure 5.4 and the assumptions
1It i“ rzlzvvnt to mzntion thvt I hvvz informvtion not only of thz populvrity of zvxh xhvnnzl wut
vl“o on xhvnnzl oryzringC
FEG gesourxe vny energy effixient network
made above, ceighwours = 2 and i = vlwvys). As shown before, with this scheme
user experience is improved. In scenario 2 selective joining is used, so not all TV
channels need to be distributed to the DSLAM. Fewer channels joined by the DSLAM
represents bandwidth savings which are translated in energy savings, as reported in
this chapter. As predictive pre-joining is not used, all switching requests experience
the normal IPTV delay. Finally, the results from scenario 3 illustrate the compromise
between operational efficiency (i.e., bandwidth and energy savings) and user quality
of experience. By using the two schemes proposed in chapters 5 and 6 it is possible
not only to improve user experience but also to increase network efficiency (although
less significantly than in scenario 2). The reason is twofold. As more channels are dis-
tributed to each STB (the requested channel plus the neighbours), the number of active
channels that need to be joined by the DSLAM naturally increases, when compared
to scenario 2. However, this increase is not very significant. The reason may reside
in the neighbouring channels of popular channels also being popular. That being the
case, there is a good probability that these channels are already amongst the active
channels joined by the DSLAM. If the period the neighbouring channels are sent con-
currently with the requested channel is finite (i.e., if i ̸= vlwvys), then the percentage
of switching requests that experience no delay is reduced (again, I refer the reader to
Figure 5.4). As in this case fewer channels will be distributed from the DSLAM to
the households one also expects the number of TV channels joined by the DSLAM to
decrease. This situation is depicted as the green arrow in the figure (note that this
decrease is not necessarily linear, as the arrow suggests).
KCL Conxlusions
Delivering TV streams in an IP network consumes a significant amount of resources.
As the number of TV channels increases and the quality of the streams improves (with
the resulting increase of its bandwidth requisites), resource and energy efficiency will
increasingly become a concern. IPTV service providers will therefore need to reconsider
their IPTV distribution networks. Fortunately, the majority of users tend to enjoy
the same TV channels: 90% of all TV viewing is restricted to a small selection of
channels [42, 161]. In this chapter I showed that IPTV providers should take advantage
of this fact. Instead of multicasting all TV channels continuously everywhere, IPTV
networks should judiciously choose fihixh TV channels to distribute fihzrz, at any one
time. In other words, they should move from their static multicast distribution schemes.
I call the method I have proposed to achieve this goal, selective joining. Contrary
6CL Xonxlusions FEH
to static multicast solutions, where network nodes join all TV multicast groups, in this
scheme the nodes join only a selection of channels. Namely, the active TV channels
(those for which there is at least one viewer connected to that node) plus a small subset
of the inactive ones (those for which that particular node has no viewers). I evaluated
selective joining by performing a trace-driven analysis using the dataset described in
Chapter 4. I contrasted the bandwidth savings achieved with the number of requests
affected, and concluded that a tradeoff is possible. Bandwidth can be reduced signifi-
cantly by distributing less TV channels in the network, without compromising service
quality, i.e., affecting only a very small percentage of channel switching requests.
A power consumption model was also developed to assess how these bandwidth sav-
ings translate into energy savings. The main conclusions were that despite nowadays
energy savings not being significant, in a plausible medium term scenario the energy
advantage of using such dynamic multicast distribution scheme becomes evident. And
as network equipment evolves to having more energy-proportional power consumption
profiles, using the selective joining scheme increases its relative advantage further when
compared with static multicast.
The first technique I proposed in this dissertation to increase the resource and energy
efficiency of an IPTV network was based on a simple paradigm: “avoid waste!” [167].
The technique I present in the next chapter is based on a different paradigm: the intro-
duction of energy-efficient optical switching technologies in these distribution networks.

Chvpter L
dptixvl wypvss of populvr ik
xhvnnels
In the previous chapter I proposed a technique to increase the resource and energy effi-
ciency of IPTV distribution networks. This technique was based on a simple paradigm:
avoiding waste. The technique I propose in this chapter is based on a different paradigm:
introducing optical switching in the network. The rationale for this proposal is the fact
that optical switching techniques are more energy-efficient than their electronic coun-
terpart. In particular, I assess the opportunities for performing optixvl wypv““ in IPTV
networks. With optical bypass, traffic not destined for a given network node is not
processed electronically by that node. This traffic is all-optically switched, i.e., it is
switched at the optical layer and is therefore not processed by the IP layer. By avoiding
electronic processing and performing optical switching instead, energy savings are to
be expected.
In this chapter I propose a novel znzrgy and rz“ourxz friendly protocol for core
optical IPTV networks. The fundamental concept is to blend electronic routing —
switching at the IP layer — and optical switching — switching at the optical layer.
The objective is to glue the low-power consumption advantage of circuit-switched all-
optical nodes with the superior bandwidth-efficiency of packet-switched IP networks.
The former assures the energy-friendliness of the scheme, whereas the latter guarantees
its resource-friendliness. The main idea is to optically switch popular TV channels while
still processing electronically the unpopular ones. Popular TV channels are watched
by many, having viewers everywhere in the network, at any time. Even considering a
semi-dynamic multicast network, as proposed in the previous chapter, these channels
have to be distributed continuously everywhere. This type of long-lived flow is the
FE6 dptixvl wypvss of populvr ik xhvnnels
perfect target to optically bypass the core network nodes. The unpopular channels
have less viewers, and hence do not need to be distributed continuously everywhere.
For bandwidth efficiency reasons, these channels are switched at the IP layer, to allow
their quick removal from or insertion to the multicast network as needed.
By analysing the dataset described in Chapter 4, I assess the opportunities for
optical bypass when using the proposed protocol in a real IPTV network. I observe
that 50% of the TV traffic can be optically switched at the network core. Additionally,
I demonstrate that the protocol does not impose significant control overhead to the
network. As its update interval can be long, it is possible to guarantee a low overhead
without compromising performance.
As the main objective of the proposed scheme is to reduce energy consumption, I
analyse its impact in this respect. The main conclusion is that with the introduction
of optical bypass the energy advantage increases further and quite significantly when
compared with the scheme proposed in the previous chapter.
LCF Introyuxtion
To guarantee the quality of experience its users demand, current IPTV networks dis-
tribute all TV channels continuously everywhere. I have shown in Chapter 6 that this is
a wasteful use of network resources and that it has an impact in energy consumption.
Fortunately, a majority of users enjoy the same TV channels, allowing the distribu-
tion of only a selected set of TV channels without impacting the expected quality of
experience significantly. Energy can therefore be saved by avoiding waste, as I have
demonstrated before.
With the goal of reducing energy consumption even further, in this chapter I con-
sider the introduction of energy-friendly optical switching techniques in the core of
optical IPTV networks. While in the previous chapter I considered only the IP layer,
being agnostic to the layers below, in this chapter I consider the particular case of opti-
cal IP networks, and therefore also the optical layer. In legacy (or first generation) core
optical networks, optics was essentially used for transmission and simply to provide
capacity [162]. All the switching and other intelligent network functions were handled
by electronics. These networks were therefore optical-electrical-optical (OEO) based,
with all traffic routed to a node being converted to the electric domain, regardless of
weather or not the traffic was destined for that node [175]. The second generation
optical networks include routing and switching at the optical layer [162]. The most
significant development of this new type of networks is the advent of optical bypass,
LCF Introyuxtion FEL
where traffic transiting a node can remain in the optical domain, instead of performing
energy-costly OEO conversions [175].
With the introduction of optical bypass capabilities in the IP network, traffic not
destined for a given IP router is placed onto a WDM wavelength that is not processed
by that router. Instead, this traffic is all-optically switched. The use of this technique
allows some work to shift from electronic routers to optical switches, which is seen as
an important strategy for managing the growth of network power consumption in the
future [14, 180, 220]. Besides reducing electronic processing in routers, the potential
for energy savings arises from the switching energy required by an all optical cross
connect being orders of magnitude below that of electronic routers [14, 205]. Due to
the circuit-switching nature of optical networks [162], however, only longBlivzy flofi“
can be considered realistic targets for optical bypass. Conveniently, some IPTV traffic
is in this category. Some TV channels are very popular, having viewers everywhere
in the network, at any particular time. Optically switching such long-lived flows can
therefore be advantageous energy-wise. Other less popular and niche channels have
periods without any viewers in particular locations, so it is wasteful to distribute them
continuously everywhere. The dynamic nature of electronic packet-switching nodes is
therefore ideal to switch this type of traffic. This guarantees the network is bandwidth
efficient, by allowing these TV channels to be quickly removed from or added to the
network as needed.
Considering the above, in this chapter I propose a hywriy protocol to be used in
the core of IPTV distribution networks, blending electronic routing with all-optical
switching. Its “hybrid” nature comes from the assumption that the network core is
composed of hybrid nodes, each including a WDM optical cross connect (OXC) and
a multicast-enabled IP router, as illustrated in Figure 7.1. The inclusion of the OXC
between the input ports and the router allows optical bypass to be performed. The main
idea of the scheme is for popular TV channels to be all-optically switched (switched at
the optical layer), while the rest are electronically routed (switched at the IP layer).
The network distributes the two different groups of channels in two (disjoint) sets of
wavelengths. The wavelengths from one set optically bypass the nodes, whereas the
other wavelengths are sent to the routers for processing.
I evaluate the proposed protocol by means of a trace-driven analysis of the dataset
described in Chapter 4. First, I demonstrate that the protocol is scalable as its update
interval can be long. This implies that the control overhead is small. Such result
was expected as channel popularity is relatively stable over short time frames [161].
Afterwards, I show that half of all TV traffic can be optically switched in the network
FEM dptixvl wypvss of populvr ik xhvnnels
core IP network
IPTV
head-end
core
network
router
OXC
Figure 7.1: Core network topology considered
core without decreasing bandwidth efficiency.
To understand the impact of this protocol on energy consumption I develop a power
consumption model for the hybrid node considered. The power consumption model for
the router is the one based on real measurements [184] used in the previous chapter. The
model for the optical layer components is based on specifications from manufacturers
and on real measurements. The main conclusion of the analysis is that the introduction
of optical bypass further increases the energy savings already achieved by using the
selective joining scheme proposed in the previous chapter. At normal traffic loads (less
than 30%), the power savings considering current and futuristic scenarios jump from
10% to 15% using selective joining only to more than 40% if one considers optical
bypass.
But it is possible to increase energy efficiency even further. Considering the baseline
traffic above, distributing all TV channels optically (i.e., with all IPTV traffic optically
bypassing the core nodes), instead of only the popular channels, would increase power
savings to around 60%. However, this comes with an increase of bandwidth inefficien-
cies, as all channels have to be distributed. By using the hybrid scheme I propose in
this chapter bandwidth can be reduced by around 20%. The growing importance of
niche channels and the expected increase in the quantity and quality of TV channels
already discussed in the previous chapter argue in favour of such hybrid rz“ourxz and
znzrgy efficient schemes.
The rest of this chapter is organised as follows. In Section 7.2 I explain how optical
bypass can be used to reduce the energy footprint of IP networks. Then I describe the
LCG ihe use of optixvl wypvss to svve energy FEN
protocol proposed in this chapter in Section 7.3. I detail the methodology used in the
analysis in Section 7.4, and evaluate the use of this protocol in Section 7.5. I analyse
its impact on energy consumption in Section 7.6, discuss the advantages of such hybrid
scheme when compared to pure all optical distribution in Section 7.7, and close this
chapter in Section 7.8.
LCG ihe use of optixvl wypvss to svve energy
An optical IP network can be seen as being made up of two layers, the IP layer and the
optical layer [85]. This is shown in Figure 7.2. In the IP layer, a core IP router connects
to an optical switching node (an Optical Cross Connect, OXC, in the figure) via short-
reach interfaces and aggregates data traffic from low-end access routers. The optical
layer provides capacity for the communication between IP routers. The OXCs are in-
terconnected with optical fibre links, each usually containing several wavelengths using
WDM technology. Associated with each fibre, a pair of wavelength MUX/DEMUX are
deployed to multiplex and demultiplex wavelengths. Associated with each wavelength,
one transponder is connected to the router to perform OEO conversions and transmit
the data [180].
core
router
low end
routers
OEO
converters
WDM links
IP
layer
optical
layer
lighpath 1 (no bypass)
lighpath 2 (optical bypass)
OXC
DEMUX MUX
Figure 7.2: Optical network employing optical bypass techniques
In the first generation of optical networks, all the lightpaths1 incident to a node
had to be terminated, i.e., all the data carried by the lightpaths would be processed
and forwarded by IP routers. This is represented in the figure by lightpath 1. The red
1gzxvll from Xhvptzr H =hzxtion HCI) thvt v lightpvth i“ vn optixvl pointBtoBpoint xonnzxtion from
v “ourxz to v yz“tinvtionC
FFE dptixvl wypvss of populvr ik xhvnnels
wavelength is OEO converted at each node. In contrast, the new generation of optical
networks includes elements such as the OXCs which allow some lightpaths to bypass
the node. This approach allows IP traffic whose destination is not the intermediate
node to directly bypass the intermediate router via a cut-through lightpath. This is
represented by lightpath 2. The green wavelength bypasses all nodes.
Several researchers have pointed out recently that optical bypass technology is one
important method to reduce the power consumption of IP networks [14, 220]. Shen and
Tucher [180], Hou zt vlC [104] and others have indeed proposed using optical bypass to
reduce the energy consumption of IP over WDM networks. This technique can save
energy because it can reduce the total number of active IP router ports, and these play a
major role in the total energy consumption of an optical IP network [180]. Introducing
optical bypass results in the possibility of some ports and even whole linecards being
turned off. Also, as the OEO converters consume a significant amount of energy, by
reducing their use the node’s energy footprint is further reduced. Shifting traffic from
power hungry routers to low power optical switches by means of optical bypass is
therefore an effective technique to save energy in optical networks.
LCH erotoxol for optixvl wypvss in Ieik
In modern IP networks most packets transit multiple core routers [142]. These packets
are fully processed at each intermediate node, with its headers inspected and forwarding
lookup being performed. These are very energy-consuming tasks [15]. I propose in this
chapter some of the core network’s IPTV traffic to optically bypass the routers, thus
reducing such electronic processing. As I already mentioned in this dissertation, some
TV channels are very popular [42, 161] having viewers everywhere in the network. My
proposal is to distribute these popular TV channels in a specific set of wavelengths
that are all-optically switched in the intermediate nodes. The rest of the TV channels
are distributed on a disjoint set of wavelengths that is sent to the routers for electronic
processing. As these channels are distributed only upon request, in a particular moment
a TV channel without viewers is not distributed. I showed in the previous chapter that
not distributing channels without viewers in the network core (i.e., by using the scheme
evaluated in section 6.4 with fkactfse\pet\pfwe set to zero) does not compromise user
experience significantly.
I assume each core network node is a hybrid node as in Figure 7.1. Each node
includes a multicast-capable optical cross connect (OXC) where optical bypass can
be performed, and a multicast-enabled IP router. I further assume these nodes are
LCH erotoxol for optixvl wypvss in Ieik FFF
GMPLS-capable. As explained in Chapter 3 (Section 3.4), a unified control plane such
as GMPLS allows the integration of optical circuit-switching techniques with electronic
packet-switching.
Vlgorithm F Processing at the IPTV source
FO while true yo
GO sleep(∆)
HO send to core-reg nodes(ACTIVE CHANNELS REQUEST)
{Wait until all requests are received...}
IO Ceop← ALL TV CHANNELS
JO Cconeop← ∅
KO for i = 1 to NUMBER OF NODES yo
LO Ceop← Ceop ∩VxtiveCh[i]
MO eny for
{Ceop now includes all popular TV channels}
NO for i = 1 to NUMBER OF NODES yo
FEO Cconeop← Cconeop ∪ (VxtiveCh[i] R∈ Ceop)
FFO eny for
{Cconeop now includes the other TV channels with viewers}
FGO o ← [Wavelengths filled with Ceop channels]
FHO e ← [Wavelengths filled with Cconeop channels]
FIO send to all nodes(SWITCHING CHANGE REQUEST, o, e)
FJO eny while
Vlgorithm G Processing at each core-regional node
FO while true yo
GO MESSAGE = msg rcv from source()
HO if MESSAGE == ACTIVE CHANNELS REQUEST then
IO VxtiveCh← get(bxvstFwyivwle)
JO send to source(VxtiveCh)
KO eny if
LO eny while
The protocol for optical bypass in IPTV networks proposed here consists of three
algorithms. Algorithm 1 runs at the IPTV source, algorithm 2 runs at core-regional
nodes (I refer the reader to the reference architecture in Figure 2.5), and algorithm
3 runs at the core nodes (including core-regional ones). The details of the proposed
protocol follows:
1. After a specified time interval, ∆ , the source transmits a message requesting all
hybrid core-regional nodes to submit their active channels (algorithm 1, lines 2-3).
FFG dptixvl wypvss of populvr ik xhvnnels
Vlgorithm H Processing at each core node
FO while true yo
GO MESSAGE = msg rcv from source()
HO if MESSAGE == SWITCHING CHANGE REQUEST then
IO for vll  ∈ o yo
JO switch optically()
KO eny for
{Wavelengths in the set o are optically bypassed}
LO for vll  ∈ e yo
MO route electronically()
NO eny for
{Wavelengths in the set e are sent to the router}
FEO eny if
FFO eny while
Recall that an vxtivz xhvnnzl is a channel for which there is at least one viewer.
This message sent by the source serves as a trigger for all core-regional routers to
send this information back to the source as soon as possible. Considering that all
nodes are GMPLS-capable, this information can be sent as an RSVP-TE Notify
message, for example. RSVP-TE Notify messages were added to RSVP-TE1 to
provide general event notification to nonadjacent nodes [154].
2. Each regional-core node then sends information on its vxtivz xhvnnzl“ to the IPTV
source. As the active channels are those being distributed by the regional-core
router to its region, the multicast forwarding table of this router contains a line
with their multicast group addresses and the interfaces used to forward packets
to2. The information requested can thus be easily retrieved and sent back to the
source (algorithm 2, lines 3-6). Again, an RSVP-TE Notify message can be used
for this purpose.
3. Once the source receives these sets from all routers, it checks which TV channels
should be optically switched (the popular ones), and which should be electron-
ically routed (the remainder channels with viewers). The popular channels are
those which have viewers everywhere. Their multicast group addresses are present
in the multicast forwarding tables of every core-regional router. The intersection
1V“ it“ nvmz impliz“A thz gz“ourxz gz“zrvvtion erotoxol B irvffix Enginzzring =ghkeBiE) i“ vn
zxtzn“ion of thz rz“ourxz rz“zrvvtion protoxol =ghke) for trvffix znginzzringA vny i“ u“zy v“ pvrt of thz
Gbeah xontrol plvnz for thi“ purpo“zC
2cotz thvt thz multixv“t “tvtz of vll vxtivz xhvnnzl“ i“ mvintvinzy in thz forfivrying tvwlz of thz
xorzBrzgionvl routzrA inxluying tho“z xhvnnzl“ thvt vrz wzing vllBoptixvlly “fiitxhzyC
LCI bethoyology FFH
of all sets received by the source thus results in a new set with the list of popular
channels1 (algorithm 1, lines 6-8). The union of the active channels of each set
which are not popular results in a new set with the non-popular TV channels
(algorithm 1, lines 9-11).
4. The TV channels are distributed, from the source, in two distinct sets of wave-
lengths: o and e. The popular channels are distributed using c different
wavelengths: o = c × . The others are sent in a disjoint set of b different
wavelengths: e = b × . The number of wavelengths in each set depends on
the number of TV channels and its bit rate, and on the capacity of each wave-
length. The IPTV source decides the composition of each set of wavelengths and
informs all core nodes of its decision (algorithm 1, lines 12-14). This information
can be sent in the form of an RSVP-TE PATH message. This is one of the mes-
sages used to allocate resources in the network. In multicast scenarios, only one
PATH message needs to be sent to multiple receivers, thus conserving network
bandwidth.
5. Each core node then sets up its switching state to optically switch the o group
(these wavelengths will therefore optically bypass the routers), and electronically
route the e group (algorithm 3, lines 3-10).
LCI bethoyology
The scheme proposed in this chapter is evaluated by means of a trace-driven analysis.
The IPTV trace detailed in Chapter 4 is used as input to the analysis performed. As I
mentioned before, I restrict the analysis of the proposed scheme to the optical network
core, as this is the only location where it is realistic to assume the presence of OXC
equipment in the medium-term. Recall that in Chapter 6 I parsed the IPTV trace data
with the objective of creating a single time-ordered trace file that includes all switching
events sent to all DSLAMs in a “pzxifix rzgion. This allows the evaluation of this scheme
at the core-regional router level, which is my intention here.
To evaluate the proposed scheme I developed a Python script that checks each line
of the input trace, to obtain each switching event that occurs in that specific region. In
a similar manner to what was done in the previous chapter, for every switching event
I record the set of active channels, actfse\ceakkeip. For each core-regional router
1I vm vwu“ing thz tzrm qpopulvr7 in thi“ xhvptzrC If onz ik xhvnnzl hv“ v “inglz vizfizr in zvxh
rzgion thzn it i“ inxluyzy in thz populvr “ztC I u“z thi“ tzrm to zv“z thz unyzr“tvnying of thz “xhzmzC
FFI dptixvl wypvss of populvr ik xhvnnels
I maintain one such structure. The intersection of all sets, at time t, is the set of
popular channels, mlm, at time t. These channels have at least one viewer per region.
The reunion of all channels that, at time t, are at least in one actfse\ceakkeip set
but are not in the mlm set is the set of non-popular channels, rkmlm, at time t. These
channels have at least one viewer in the network, but there are regions where they have
no viewers. The channels that are not in the mlm nor in the rkmlm sets are included
in the kl\sfet set. These channels have no viewers anywhere in the network. By
running this script I am thus able to know, with the precision of one second for the
trace duration (recall that the trace has one second precision), the number of channels
with users everywhere (popular), somewhere (unpopular), and nowhere (no viewers).
Figure 7.3 illustrates the proposed methodology with a simple example. I assume
the network distributes only five channels, numbered from 1 to 5. At 12:41:36am an
RM message for channel 1 is received in node x1’s region from the STB with IP address
.-.41.26.65. In regions x2 and x3 RM messages for channel 2 are sent at exactly the
same time. These two channels are hence included in the rkmlm set while the other
three channels remain in the kl\sfet set. The set mlm remains empty as there are no
channels with viewers everywhere. Nine seconds later node x1 receives a join message
to channel 3. This channel is included in the rkmlm set. Around one minute later it is
again removed from this set, and included in the kl\sfet set, after a altk message is
sent to the same node. Finally, at 12:43:48am two UP messages are sent to channel 1
in regions x2 and x3. As this channel is now active in every region, it is removed from
the rkmlm set and included in the set mlm.
LC5 Evvluvtion
As explained before, the proposed protocol is evaluated by performing a trace-driven
analysis on the IPTV dataset. All results I present in this chapter arise from the
analysis of the whole data set (6 months, 255 thousand users). The evaluation is
threefold. First, I investigate the scalability of the protocol. Second, I analyse the
opportunities for optical bypass when running the proposed protocol in the network
under study. Finally, in the next section I analyse the impact the use of this protocol
has in power consumption of the IPTV network.
LC5CF hxvlvwility
For a network protocol to be scalable it is important that it does not impose a signifi-
cant processing overhead to the network nodes and that it does not add a great amount
LC5 Zvvluvtion FF5
core IP network
IPTV
head-end
core
network
x
Jul 1 00:41:36|UP|1|10.74.59.98
Jul 1 00:41:44|UP|3|10.74.77.101
Jul 1 00:42:51|DOWN|3|10.74.77.101
x = {1,2,3,4,5}
x1
pop={}
unpop={1,2}
no_view={3,4,5}
pop={}
unpop={1,2,3}
no_view={4,5}
pop={}
unpop={1,2}
no_view={3,4,5}
pop={1}
unpop={2}
no_view={3,4,5}
Contents of each structure
Jul 1 00:41:36
Jul 1 00:41:44
Jul 1 00:42:51
Jul 1 00:43:48
Jul 1 00:41:36|UP|2|10.74.59.98
Jul 1 00:43:48|UP|1|10.74.80.80
x2
x3
Jul 1 00:41:36|UP|2|10.74.59.98
Jul 1 00:43:48|UP|1|10.74.80.80
Figure 7.3: Proposed methodology
of signalling traffic to the network. By guaranteeing a relatively long update interval
for the control information (the ∆ variable in the proposed protocol) it is possible to
guarantee a low overhead to the nodes and to the network as a whole. On the other
hand, to assure the best performance it is important that the network state1 is consis-
tent with network usage (in this particular case, it should reflect channel popularity).
Having a short update interval marries with this objective.
It is known that channel popularity is relatively stable over short time frames, and
that it becomes more dynamic when longer time frames are considered [161]. Regular
updates may therefore not be needed. To attest this, I analyse the henceforth called
ik xhvnnzl xhurn rvtz in the 11 core-regional nodes of this network. I compare the
active TV channels at time  with the active channels at time  + ∆ , for different
values of ∆ . The number of channels that are different between the two sets in two
1In thi“ xontzxtA thz nztfiork “tvtz xon“i“t“ of thz fivvzlzngth “fiitxhing xonfigurvtion vt zvxh noyzA
vny thz “zt of ik xhvnnzl“ trvn“portzy in zvxh fivvzlzngth groupA o vny eC
FF6 dptixvl wypvss of populvr ik xhvnnels
consecutive periods is the TV channel churn rate. The results are shown in Figure 7.4,
for each region, and for five values of ∆ . The median of the channel churn rate over
the whole period of the trace (6 months) is presented, with the lower and upper error
bars representing the 5th- and 95th-percentile, respectively.
By analysing the results in Figure 7.4, I conclude that the churn rate is usually quite
low, particularly for values of ∆ below 1 hour. A long update interval of 15 minutes,
for instance, is a good compromise. It does not represents a significant overhead to
the network, while at the same time guarantees that the network state changes with
channel popularity dynamics.
LC5CG dpportunities for optixvl wypvss
The protocol proposed in this chapter divides the TV channels into three groups: the
popular channels, the unpopular channels, and the channels without viewers. The
channels from the former group optically bypass the routers. Those from the second
group are sent for the router for electronic processing. Finally, those from the latter
group are not distributed by the IPTV source. To understand the opportunities for
optical bypass in the core of the IPTV network, I need to quantify how many channels
would be included in each group at regular intervals. For this purpose, I retrieve the
number of channels in each set mlm, rkmlm, and kl\sfet periodically, for the whole
trace. I consider for the analysis an update interval equal to 15 minutes, for the reasons
explained above. This is the periodicity with which I retrieve the number of channels in
each set. In Figure 7.5 I present the results obtained (median, 5th-, and 95th-percentile)
from the analysis of the whole dataset.
I start the analysis from the bottom. On average, one fifth of the TV channels do not
need to be distributed by the IPTV source. This is the reason why I used this number
for the analysis made in Section 6.5.2. Recall from that section that not distributing
this traffic has a negligible impact on the service (Figure 6.4). The remaining 80% TV
channels are distributed to the network core. Around 50% of the TV channels can be
optically bypassed. This means that, on average, at any one time, half of the channels
have at least one viewer in each region. The number of channels requiring electronic
processing can thus be reduced to around 30%. In the next section I investigate the
impact this has on energy consumption.
LC5 Zvvluvtion FFL
Region 11
Region 10
Region  9
Region  8
Region  7
Region  6
Region  5
Region  4
Region  3
Region  2
Region  1
Region 11
Region 10
Region  9
Region  8
Region  7
Region  6
Region  5
Region  4
Region  3
Region  2
Region  1
Region 11
Region 10
Region  9
Region  8
Region  7
Region  6
Region  5
Region  4
Region  3
Region  2
Region  1
Region 11
Region 10
Region  9
Region  8
Region  7
Region  6
Region  5
Region  4
Region  3
Region  2
Region  1
Region 11
Region 10
Region  9
Region  8
Region  7
Region  6
Region  5
Region  4
Region  3
Region  2
Region  1
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
∆
T
=
3
0
 sec
∆
T
=
1
5
 m
in
∆
T
=
1
 h
o
u
r
∆
T
=
6
 h
o
u
rs
∆
T
=
1
 d
ay
0 5 10 15 20 25 30 35
TV channel churn rate
Figure 7.4: TV channel churn rate for all eleven regions, for five values of the update
interval
FFM dptixvl wypvss of populvr ik xhvnnels
Not distributed
Electronically routed
Optically bypassed
●
●
●
20%
30%
50%
0 20 40 60 80 100
Number of TV channels
Figure 7.5: Average number of TV channels that are optically bypassed, electronically
routed and not distributed, respectively
LCK Impvxt on energy xonsumption
After understanding that by using the proposed protocol there are clear opportunities
to introduce optical bypass in IPTV networks I now analyse the impact this has on
energy consumption. By employing this technique energy savings are expected for two
reasons:
1. Some traffic flows (the popular TV channels) bypass some routers. This reduces
the number of bits requiring electronic processing, thus avoiding energy-expensive
OEO conversions, buffering, and forwarding table lookups. The work is shifted to
optical switches, which are at least two orders of magnitude more energy efficient
when compared to its electronic equivalent [14].
2. As TV channels without viewers are not distributed, network load is reduced and
even less bits require electronic processing in the routers.
LCKCF helextive joining in xore optixvl networks
Before presenting results from using the proposed scheme, it is important to return to
the scheme proposed in the previous chapter, selective joining, considering an optical
network scenario. In Chapter 6 I focused the analysis on power consumption of IP
routers only. As now I consider a core optical network, I also need to include the power
consumption of the optical layer components. This will allow a fair comparison with
the proposal made in this chapter.
In optical networks, associated with each wavelength (port) is a transponder (OEO
converter), as was shown in Figure 7.2. The transponder interfaces the router to a fibre
optic cable. Its main function is to perform the required OEO conversions. Considering
LC6 Impvxt on energy xonsumption FFN
this, the power consumption model presented as Equation 6.1 is now updated, as in
Equation 7.1.
e = ech +Kiei +
L∑
i20
eli (7.1)
As can be seen, the only difference from equation 6.1 is the inclusion of the power
consumption of the transponders. In this equation, Ki is the number of transponders
(one per port) and ei is the power per transponder. Every time a new port needs to
be turned on, a new transponder is also activated. I assume the power consumption
for each transponder to be 73 W, based on Alcatel-Lucent WaveStar OLS 1.6T ultra-
long-haul systems [6]. This figure has been used in recent related work [104, 180].
In accordance to the results presented in the previous section (Figure 7.5), I assume
that only 20% of the channels are not distributed to the core. The results I present in
Figure 7.6 thus correspond to a reduction of IPTV traffic in the network core to 80%.
0%
5%
10%
15%
20%
25%
linecard turned on
in baseline scenario
transponder activated in
selective joining scenario
25% 35% 45% 55% 65% 75%
Baseline traffic load (%)
P
o
w
er
 s
av
in
g
s 
(%
)
3kUHD_ep (IP + opt) 3kUHD_ep (IP only)
3kUHD (IP + opt) 3kUHD (IP only)
700HD (IP + opt) 700HD (IP only)
150SD (IP + opt) 150SD (IP only)
Figure 7.6: Power savings of using the selective joining scheme considering an optical
IP network, as a function of the baseline traffic load in the node (core network)
As can be seen by comparing this plot with the one presented in the previous
chapter, in Figure 6.81, the results change significantly. The main reason is the fact that
the transponders are power hungry equipment. This results in an increased advantage
in using the selective joining scheme in some scenarios, as reducing traffic load decreases
the number of active transponders. It is particularly relevant to mention scenario 4--EA,
1ihz yv“hzy linz“ in [igurz LCK vrz thz “vmz v“ thz yv“hzy linz“ in [igurz KCMC
FGE dptixvl wypvss of populvr ik xhvnnels
which is typical in current networks (recall that this scenario is based on AT&T’s IPTV
service offering [160]). The use of the scheme proposed in the previous chapter increases
the power savings to around 10% in normal traffic conditions. Similarly to Figures 6.7
and 6.8 in Chapter 6, the peaks in Figure 7.6, evident in both the .2-PA and 4--EA
scenarios in this case, represent transition points. In the previous chapter, the most
pronounced bumps represented new linecards being turned on in the baseline scenario.
While the traffic load did not increase over those transition points the proposed scheme
presented a higher-than-average power saving advantage. The same occurs in Figure
7.6. However, in this case the peaks represent the addition of another active transponder
in the baseline scenario. As this component consumes more power than a linecard, the
peaks are more pronounced when compared with those from the previous chapter. They
also occur more frequently because an active transponder is activated every time a new
port is turned on.
LCKCG Energy xonsumption moyel of the hywriy noyes
To be able to quantify the energy savings achieved by introducing optical bypass in an
optical IPTV network, in this section I build a power consumption model of the hybrid
node considered in this chapter. Such node is depicted in Figure 7.7.
OXC
OEO
converters
core router
Figure 7.7: Hybrid node
Three factors affect the power consumption of an hybrid node:
1. The power consumption of the router.
2. The power consumption of the OXC.
3. The power consumption of the OEO converters (transponders).
Note that in this analysis I do not consider the power consumption of other optical
equipment that is necessary in an optical network, such as the optical amplifiers, multi-
plexers and demultiplexers. I consider switching equipment and OEO converters only.
LC6 Impvxt on energy xonsumption FGF
Previous work [180, 220] as shown that switching equipment and transponders (OEO
converters) are the main contributors for power consumption of optical IP networks
(responsible for over 97% of total power consumption according to [180]). Based on
the three variables above, I use the following model for the power consumption e of a
hybrid node:
e = eR + eOXX + eOEO (7.2)
In equation 7.2 eR is the power consumption of the router, eOXX is the power con-
sumption of the optical cross connect, and eOEO is the power consumption of the OEO
converters (transponders). For eR I use the model based on real measurements [184]
developed in Chapter 6 (Section 6.5.1). The power consumption of the OXC is given
by equation 7.3.
eOXX = Kopeop (7.3)
In this equation, Kop is the number of input/output optical switch ports and eop is
the power per input/output switch port. I assume the OXC switching fabric is realised
using micro-electro-mechanical systems (MEMS) [76]. In a MEMS optical switch, a
micro-mirror is used to reflect a light beam. The direction in which the light beam is
reflected can be changed by rotating the mirror to different angles, allowing the input
light to be connected to any output port. These MEMS have switching times of the
order of milliseconds or hundreds of microseconds and for this reason can be used only
for slow switching (i.e., circuit switching). For faster switching Semiconductor Opti-
cal Amplifiers (SOAs) could be used. But as MEMs consume less power [7], and as
the OXC is not to be used for fast switching, MEMS are the option here. I assume
3D-MEMS [215] in particular. The power per input/output switch port of the OXC
corresponds to the power consumption for its continuous control, which is equal to
107 mW per input/output port. This value is based on the power consumption of the
MEMS controller circuitry of an 80 × 80 3D-MEMS switch implementation, reported
in [215]. I am therefore assuming power consumption is proportional to the number of
active input/output ports1. The experimental figure and this assumption were consid-
ered in previous related work [7, 76] and are also in agreement with studies from other
researchers [14, 193].
1If I v““umz vn onDoff wzhvviourA iCzCA v “fiitxh xon“uming it“ MCJ l of totvl pofizr inyzpznyzntly of
thz numwzr of vxtivz port“A vll rz“ult“ I prz“znt in thi“ xhvptzr xhvngz wy lz““ thvn F:C ihi“ “tzm“ from
thz fvxt thvt thz dmX i“ thz noyz xomponznt fiith thz lofiz“t pofizr xon“umption wy v gooy mvrginA
in vny xv“zC
FGG dptixvl wypvss of populvr ik xhvnnels
Finally, the power consumption of the OEO converters is given by equation 7.4.
eOEO = Kiei (7.4)
In this equation, Ki is the number of transponders (one per wavelength that con-
nects to the router) and ei is the power per transponder. As in the previous subsection,
I assume the power consumption for each transponder to be equal to 73 W.
LCKCH gesults
I now analyse how the introduction of optical bypass techniques in the IPTV network
translate into energy savings. I consider the same three scenarios as in Chapter 6:
.2-PA, an IPTV service offering of 150 SDTV channels; 4--EA, 700 HDTV channels;
and 0hREA, 3000 UHDTV channels. For the router model I make the same assumptions
as in Chapter 6. For the first scenario, I assume a router with 4 linecards with 4x1Gbps
Ethernet ports each, as in Figure 6.6. For the other scenarios I just scale up the model
by increasing the number and changing the type of linecards. This implies that each
wavelength can carry 1Gbps in the first two scenarios, but it scales to 40 Gbps in the
third. Note that in this scheme two sets of wavelengths are needed: one for the traffic
that optically bypasses the routers, and another for the rest. This is considered in the
analysis to calculate the number of active OXC ports. The number of active OEO
converters is equal to the number of active ports in the router.
In accordance to the results presented in Figure 7.5, I assume that 50% of the
IPTV traffic optically bypasses the routers, 30% is sent to the router for electronic
processing, and 20% of the TV channels are not distributed. Considering this, the
power savings for all three scenarios (plus the 0hREA\em scenario, which is the same as
0hREA but considering routers with an energy-proportional power consumption profile)
are presented in Figure 7.8. The dashed lines represent the results from using selective
joining only. The solid lines represent the power savings using the optical bypass
protocol proposed in the current chapter.
When compared with the selective joining scheme proposed in the previous chapter,
the introduction of optical bypass in the optical IPTV network core increases power
savings substantially. At baseline traffic loads of around 30%, the power savings increase
from 10% to 15% to over 40%. Considering EP routers, power consumption is halved. I
conclude that the use of this technique is very effective in reducing power consumption,
including in current IPTV service scenarios (such as 4--EA).
LCL DisxussionO on the vvlue of elextronixs FGH
0%
10%
20%
30%
40%
50%
60%
25% 35% 45% 55% 65% 75%
Baseline traffic load (%)
P
o
w
er
 s
av
in
g
s 
(%
)
3kUHD_ep (bypass) 3kUHD_ep (no bypass)
3kUHD (bypass) 3kUHD (no bypass)
700HD (bypass) 700HD (no bypass)
150SD (bypass) 150SD (no bypass)
Figure 7.8: Power savings achieved by optically bypassing popular TV channels in the
network core, as a function of the baseline traffic load in the node
LCL DisxussionO on the vvlue of elextronixs
In the previous section I showed how optically switching populvr IPTV traffic reduces
power consumption significantly. How about optically switching vll IPTV traffic? To
answer this question, I invite the reader to look at Figure 7.9. This graph shows the
result of optically switching all IPTV traffic in the network core (solid lines), against
optically switching only the popular TV channels (dashed lines).
As can be observed, by optically switching all IPTV traffic the power savings in-
crease even further. Considering a baseline traffic load of 25%, in the 4--EA, 0hREA and
0hREA\em scenarios an additional 20% power saving is achievable by all TV channels
bypassing the routers.
So why not moving completely to optics in the future? In a scenario where all
IPTV traffic is optically bypassed, to guarantee their availability for IPTV users, all
TV channels need to be distributed continuously in the network core. This is because
OXCs allow slow switching only1. The advantage of maintaining the electronic routing
option is that, contrary to circuit-switched optical networks, with electronic routing it
is possible not to distribute all TV channels. This added capability increases bandwidth
1Wy “lofi “fiitxhing I mzvn “fiitxhing tzxhnologiz“ fiith “pzzy“ on v milli“zxony rvngzC [or in“tvnxzA
thz HDBbEbh “fiitxhz“ xon“iyzrzy in thi“ xhvptzr hvvz v “fiitxhing timz of vrouny FEm“ pGFMrA “o vn
optixvl xro““ xonnzxt wv“zy on thi“ tzxhnology i“ “lofi to rzxonfigurzC ihi“ i“ in xontrv“t to fv“t “fiitxhing
=nvno“zxony rzgimz) v“ rzquirzy for pvxkzt “fiitxhingA for fihixh no mvturz optixvl tzxhnology i“ yzt
vvvilvwlz pGFMrC
FGI dptixvl wypvss of populvr ik xhvnnels
0%
20%
40%
60%
80%
25% 35% 45% 55% 65% 75%
Baseline traffic load (%)
P
o
w
er
 s
av
in
g
s 
(%
)
3kUHD_ep (all) 3kUHD_ep (pop)
3kUHD (all) 3kUHD (pop)
700HD (all) 700HD (pop)
150SD (all) 150SD (pop)
Figure 7.9: Power savings achieved by optically bypassing all TV channels in the net-
work core (compared to popular TV channels only), as a function of the baseline traffic
load in the node
efficiency. As explained in the introduction, with the increased popularity of narrow-
casting services and niche channels, the number of unpopular channels (as defined in
this chapter) may plausibly increase to the several hundreds or thousands in the near
future. This trend offers an important argument for the maintenance of electronic rout-
ing as an option. A hybrid scheme as the one proposed in this chapter therefore offer a
compromise between znzrgy and rz“ourxz efficiency IPTV service providers may want
to consider.
LCM Conxlusions
In this chapter, I considered the introduction of energy-friendly optical technologies to
reduce the energy consumption of IPTV distribution networks. I proposed an energy
and resource-friendly protocol for the IPTV network core, blending electronic routing
with all-optical switching. The main idea is to optically switch popular TV channels.
This IPTV traffic bypasses the routers and therefore does not require any electronic
processing (it is switched at the optical layer). The rest of the channels are sent to the
routers for electronic processing (to be switched at the IP layer).
By analysing the IPTV dataset described in Chapter 4, I observed that by using
the proposed protocol it is possible to switch 50% of the IPTV traffic all-optically. The
energy savings obtained from optically bypassing this traffic are substantial. When
LCM Xonxlusions FG5
compared with the selective joining scheme proposed in the previous chapter, the power
savings in the network increase from 10% to 15% to over 40% under normal load
conditions. The scheme is also bandwidth efficient as channels without viewers are not
distributed. Finally, if all IPTV traffic is optically switched, instead of only the popular
TV channels, the power savings increase even further. However, this comes with an
increase of bandwidth inefficiency.

Chvpter M
hummvry of xontriwutions vny
future work
The closing chapter of the dissertation summarises the work upon which it is based
and its original contributions. In addition, potential avenues for future research are
proposed.
MCF hummvry of xontriwutions
In this dissertation I studied and analysed three techniques to assist IPTV providers in
the design of novel resource and energy efficient networks. These techniques addressed
two relevant technological challenges currently faced by IPTV operators. The first such
challenge is IPTV service’s high channel change delay. Synchronisation and buffering
of media streams can cause channel change delays of several seconds. The second is the
question of how to maintain an operationally cost and energy efficient network in face
of the evolution of IPTV services. Current static multicast solutions are inefficient, but
dynamic multicast solutions also bring issues related to network scalability and service
quality.
In face of these technological challenges, the first contribution of this dissertation
was an empirical analysis of a particular solution to the channel change delay problem
— predictive pre-joining of TV channels — using real IPTV usage data. In this scheme
each Set Top Box simultaneously joins additional multicast groups (TV channels) along
with the one requested by the user. If the user switches to any of these channels next,
switching latency is virtually eliminated, and user experience is improved. Previous
work on this subject used simple mathematical models to perform analytical studies
FGM hummvry of xontriwutions vny future work
or to generate synthetic data traces to evaluate these pre-joining methods. I demon-
strated in this dissertation that these models are conservative in terms of the number
of channel switches a user performs during zapping periods. They do not evidence
the true potential of predictive pre-joining solutions, and were therefore an important
motivation to perform such empirical analysis. The main conclusion of this study was
that a simple scheme where the neighbouring channels (i.e., the channels adjacent to
the requested one) are pre-joined by the Set Top Box alongside the requested channel,
during zapping periods only, eliminates zapping delay for around half of all channel
switching requests to the network. Importantly, this result is achieved with a negligible
increase of bandwidth utilisation in the access link.
The second contribution of this dissertation was related to the design and operation
of IPTV networks. Current IPTV service providers build static multicast trees for the
distribution of TV channels. This is justified to guarantee the quality of experience
required by its customers. By distributing TV channels to as close to the users as
possible, network latencies do not add significantly to the already high channel change
delay. However, as particular channels have no viewers at particular time periods,
this method is provably resource and energy inefficient. To reduce these inefficiencies,
I proposed a semi-dynamic scheme where only a selection of TV multicast groups is
distributed in the network, instead of all. This selection changes with user activity. This
method was evaluated empirically by analysing real IPTV usage data. I demonstrated
that by using the proposed scheme IPTV service providers can save a considerable
amount of bandwidth while affecting only a very small number of TV channel switching
requests. Furthermore, I also showed that although today the bandwidth savings would
have reduced impact in energy consumption, with the introduction of numerous very
high definition channels this impact will become significant.
To further increase the energy efficiency of IPTV networks, the third contribution
of this dissertation was a novel energy and resource friendly protocol for core optical
IPTV networks. The fundamental concept is to blend electronic routing and optical
switching, thus gluing the low-power consumption advantage of circuit-switched all-
optical nodes with the superior bandwidth-efficiency of packet-switched IP networks.
The main idea is to optically switch popular TV channels. These can be categorised as
long-lived flows, and are therefore perfect targets for this type of slow switching. With
the use of this protocol, popular IPTV traffic optically bypasses the network nodes,
i.e., this traffic avoids electronic processing. I evaluated this proposal empirically by
performing a trace-driven analysis using real IPTV data. The main conclusion was that
the introduction of optical switching techniques results in a quite significant increase
MCG Future yirextions FGN
in the energy efficiency of IPTV networks.
All the schemes studied in this dissertation were evaluated by means of trace-driven
analyses using a dataset from an operational IPTV service provider. It is widely ac-
cepted that a thorough evaluation using real workloads enables the assessment of future
network architectures with an increased level of confidence. This is particularly rele-
vant in research fields that have relied heavily upon hypothetical user models which are
different from the reality and can lead to incorrect estimation of system performance.
Such is the case of IPTV systems research, which favours the use of evaluation methods
as the one employed in this dissertation.
MCG Future yirextions
In this dissertation I addressed two important technological challenges currently faced
by IPTV operators: high channel change delay and network efficiency. The solutions
proposed and analysed in this dissertation mitigate part of these problems. But many
more persist. As such, I close this dissertation by suggesting possible directions for
future research on these topics.
MCGCF Improving xhvnnel xhvnge user experienxe
Most commercial solutions to the channel change delay problem attempt to ensure that
an STB that is trying to join a new TV channel gets an auxiliary stream that starts
with an I-frame and then offers some kind of mechanism to switch over to the main
multicast stream (the boost stream solutions described in Section 3.2.2). This type of
solution requires a dedicated zapping server to transmit a unicast burst when a channel
change request is made. This is the most common fast channel change mechanism,
and is used, for example, by the Windows Media Platform [141]. As the auxiliary
stream starts with an I-frame, the zapping server maintains a delayed version of all
TV streams. The unicast stream sent to the STB after a channel request is therefore
a yzlvyzy vzr“ion of the original multicast stream. To avoid glitches, when the STB
switches to the multicast stream this one synchronises with the unicast stream, and is
delayed for play out. The multicast stream in the STB is therefore out of synch with the
original multicast stream that is being distributed in the network. In certain situations
these delays may cause discomfort to the IPTV users (for example, your neighbour
cheering a football goal before you see it), and hence is a challenge for IPTV network
operators. A possible solution to this problem which may be worth investigating is to
speed up the delayed multicast stream in order for it to catch-up the original stream.
FHE hummvry of xontriwutions vny future work
Informal subjective tests have shown that the variation of the playout speed is often
unnoticeable by users [119, 187], so it may be a solution if performed in a controlled
way. Kalman zt vlC [118, 119] have presented a similar idea in the past in order to
buffer less video data, with the objective of reducing zapping delay.
MCGCG Improving resourxe effixienxy
IPTV services are bandwidth intensive. High definition TV requires bit rates on the
tens of Mbps range, and future ultra high definition formats may increase this figure
by orders of magnitude. So resource efficiency will continue to be a challenge for IPTV
operators.
The type of solutions analysed in this dissertation to mitigate the high channel
change delay problem of IPTV services assume several TV channels are sent simul-
taneously to the Set Top Box. This increases the bandwidth requirements of access
networks. To alleviate this problem, a more efficient scheme would be to use Scalable
Video Coding (SVC) techniques [208] with pre-joining solutions. SVC video streams
contain one or more subset streams, or layers. A subset video stream is derived by
dropping packets from the larger video to reduce the bandwidth required for the subset
bitstream. The subset bitstream can represent a lower quality video signal, for instance.
Sending the additional TV channels in lower quality (for example, by transmitting its
base layer only) while transmitting all layers of the channel requested (to guarantee
maximum quality for this particular channel) may offer an interesting tradeoff between
switching latency and access network bandwidth cost worth investigating.
For the IP network core, constructing more efficient multicast distribution trees
is another issue that deserves investigation. To build a multicast tree, PIM-SM [73],
the most common multicast routing protocol [178], makes use of the unicast routing
protocol topology information available through the routers forwarding table. This
information, together with the group membership information, enables the construction
of shortest-path trees (SPTs) from the source or from a core node (the Rendezvous
Point in PIM). To be precise, in situations where paths are asymmetric, these are
rzvzr“z SPTs because PIM uses unicast routing shortest-paths from the receiver to the
source to build the branch of the tree from the source to the receiver [63].
In a recent paper, Xu zt vlC [213] have proved that optimal traffic engineering
(TE) can be realised using link-state routing protocols with hop-by-hop forwarding.
They presented a link-state protocol, PEFT, that provably achieves optimal traffic
engineering while retaining the simplicity of hop-by-hop forwarding. By using PIM-SM
MCG Future yirextions FHF
with such unicast protocol it is therefore possible to obtain optimal TE rzvzr“z SPTs.
But IPTV traffic consumes a very significant amount of bandwidth in the forfivry
direction, from the source to its multiple destinations.
A solution to create optimal TE forfivry SPTs could thus be the following. A
new source-initiated message would be added to PIM-SM. This message would be used
to update the multicast routing table from the source to the receiver, allowing the
construction of forward SPTs. This is a similar technique to the one used by the
multicast protocol REUNITE [188]. By using this modified PIM-SM with a unicast
protocol such as PEFT one would obtain optimal traffic engineered Shortest Path Trees
in the forward direction.
MCGCH Improving energy effixienxy
I demonstrated in this dissertation that avoiding waste and opting for low-power switch-
ing are effective techniques to improve energy efficiency in IPTV networks. These solu-
tions assumed little changes to current IPTV network architectures and topologies. An
interesting future avenue would be the research on new network architecture designs
and novel topologies with the objective of conceiving more environment-friendly IPTV
networks.
An idea for future work in this area is to investigate energy-friendly placement
strategies for PIM-SM Rendezvous Points (RPs). For scalability reasons, multicast
protocols such as PIM-SM contain the option of having a single node from which
branches of the multicast tree emanate. Scalability is obtained by the possibility of
having a single multicast tree per group as opposed to one tree per (source, group)
pair [21]. In PIM-SM this core node is called the Rendezvous Point (RP). The selection
of the RP directly affects the structure of the tree, and therefore the performance of the
network. An important problem in the construction of shared multicast trees is hence
to determine the position of the RP. The choice of the RP allows network providers to
perform traffic engineering, as in the recent work by Wang zt vlC [204]. The authors
proposed a new algorithm, based on tabu search [87], to find the optimal placement
for RP nodes. In their work, the cost to minimise to achieve this optimum is the sum
of the average throughput of all links. Instead of minimising such variable, one could
explore similar strategies that instead minimise energy consumption without negatively
impacting performance.
All these techniques reduce energy consumption, but a truly environment-friendly
IPTV network should have has its goal to reduce or eliminate the emission of green-
FHG hummvry of xontriwutions vny future work
house gases. With such purpose, Dong zt vlC [68] recently proposed a novel approach
to minimise Cd2 emissions in optical networks (considering unicast traffic). The au-
thors achieved their goal by assuming some network nodes have access to renewable
energy sources and by maximising its use in the network. Another interesting avenue
of research is to investigate similar techniques assuming IPTV multicast scenarios.
Vppenyixes

Vppenyix V
From elextronixs to optixsO
envwling texhniques
In this appendix I present some work that, despite its orthogonality to the proposal
presented in Chapter 7, is closely related. Namely, I address optical multicast, traffic
grooming, and aggregated multicast. These are important techniques for IPTV oper-
ators that want to take full advantage of the opportunities offered by the inclusion of
novel optical technologies in their networks.
VCF dptixvl multixvst
Since the seminal work by S. Deering [62], in 1989, the multicast problem has been
extensively studied in the electrical domain. More recently the research focus has
integrated multicast in the optical domain, which is the focus of this section. As
explained in Chapter 3, in an all-optical network a lightpath is an optical point-to-point
connection from a source to a destination. Switching at intermediate nodes is done at
the optical layer, so the path from source to destination is all-optical. In [174] this
concept was generalised to that of a light-tree which, unlike a lightpath, has multiple
destination nodes. An important advantage of optical multicast is signal transparency
with respect to traffic type, bit rates and protocols. In addition, Wang and Yang have
shown that the use of optical multicast leads to a significant reduction in the number
of wavelengths required in most networks, thereby increasing network efficiency [207].
Issues on optical multicast can be classified as data plane or control plane issues. At
the data plane level, the fundamental issues are the architecture of multicast-capable
optical cross connects and network topology design (in particular, the optimal place-
FH6 From elextronixs to optixsO envwling texhniques
ment of network equipment). Concerning the former, nodes with optical multicast
capability are usually implemented by using optical splitters. A light splitter has the
ability to split an input optical signal into multiple identical output optical signals.
The only difference is power reduction of the output signals. Ideally, for a light splitter
with a fan out of n, the power at each output of the splitter is 1n . The power constraints
on optical networks are therefore exacerbated by the presence of optical splitting, and
this has to be considered in the network design phase (for example, to define where to
place the optical amplifiers in the network, a problem addressed by Hamad and Kamal
in [96, 97]). Other technique to split the signal is WDM multicasting. This type of op-
tical multicast can be done by taking advantage of the non-linear nature of optical fibre
(by making clever use of Self-Phase Modulation, SPM, as in [84], or Four Wave Mixing,
FWM1, as in [78], for instance), by utilising pump modulated parametric amplifiers2, as
in [36], or by using an active vertical couplers-based optical switch3 [217]. All the node
architectures referred so far are of the Split-and-Delivery (SaD) type. A different type
of multicast node architecture was proposed by Ali and Deogun [9]: Tap-and-Continue
(TaC). Contrary to splitting the signal, as in SaD, in the TaC architecture the data
proceeds strictly along a path but intermediate nodes on the path may access the data
themselves by tapping a small fraction of the signal. This mechanism reduces splitting
loss, but still has the inherent limit on the number of times a signal can be tapped
before it loses integrity. As usual, hybrid architectures also exist. Fernandez et al. [75],
for example, proposed a novel architecture that tries to combine the advantages of both
tapping and splitting. Their GB“plitBtvpBxontinuz node is similar to a SaD-based node.
This difference is that it not only switches but also taps a fraction of the input power
to the local node. The architecture also includes a novel interconnection network which
is the key for the improved efficiency over both SaD and TaC architectures alone.
The design and optimisation of an optical network is a difficult problem, in partic-
ular due to the heterogeneity of the equipment. The cost of hardware precludes full
deployment in all nodes of optical splitters and wavelength converters. Wavelength
conversion (WC), the ability to convert an input signal received on one wavelength into
an output signal on a different wavelength, is a desirable capability for an optical node,
1heb vny [lb vrz nonBlinzvr zffzxt“ thvt vri“z in optixvl xommunixvtion “y“tzm“ yuz to thz
yzpznyznxy of thz optixvl fiwrz rzfrvxtivz inyzx on thz intzn“ity of thz vpplizy zlzxtrix fizlyC In highly
nonBlinzvr fiwrz heb wrovyzn“ thz “ignvl’“ zlzxtromvgnztix “pzxtrumC [lb oxxur“ in lDb “y“tzm“C
ihz mix of “zvzrvl fivvzlzngth“ in onz fiwrz givz“ ri“z to nzfi =u“uvlly unyz“irvwlz) “ignvl“ vt nzfi
frzquznxiz“C
2Vn optixvl vmplifizr xvpvwlz of offzring “pzxtrvlly fiiyz gvin in vny wvny of intzrz“tC
3huxh optixvl “fiitxh vllofi“ optixvl multixv“t fiithout zxxz““ optixvl “plitting lo““ yuz to thz optixvl
gvin vvvilvwlz in vxtivz vzrtixvl xouplzr “fiitxh xzll“C
VCF dptixvl multixvst FHL
as it can help improve wavelength utilisation in the network. However, the costs of
wavelength converters are still very high [222], and for that reason only a small subset
of network nodes may realistically be WC-capable. Besides this, these nodes usually
are capable of converting only from specific input wavelengths to another wavelength
within a finite waveband, so they have limited wavelength-conversion capabilities. A
multicast node is also expensive to implement due to the complexity of fabrication and
the large number of amplifiers required. For this reason, current networks not only
have sparse limited wavelength-conversion capabilities (not all nodes are WC-capable,
and the ones that are have limited conversion capabilities), but also have sparse lim-
ited multicast capabilities (not all nodes are multicast-capable, and as light splitters
have a finite fan-out they are limited on the number of outputs). If properly designed,
however, these limitations are not a serious problem. Networks with just a few power
splitters and wavelength converters have efficiencies close to that of a full WC- and
multicast-capable network, as shown by Yang and Liao’s work [214].
Besides network heterogeneity, other constraints need to be taken into account when
designing and optimising an optical network. First, each link can have multiple fibres
and multiple wavelengths per fibre. Second, light-trees that share a common physical
link cannot be assigned the same wavelength (the wavelength-clash constraint [185]).
Third, the power level of the signal on any wavelength must not degrade below a
certain lower bound [143] (namely, the sensitivity of the receivers and of the optical
amplifiers), and simultaneously should not exceed an upper bound due to the non-
linearity effects [162]. Finally, other physical-layer impairments, such as dispersion,
have to be considered [212]. The topology design problem of optimally placing the
network nodes taking into account all these constraints is therefore extremely complex.
The research on this issue has considered the optimal placement of splitters alone [8,
135] or jointly with the wavelength converters [41, 48, 216], and also assuming nodes
with both splitting and wavelength-conversion capabilities [66].
Besides the above data plane issues, optical multicast requires algorithmic support
from the control plane. On the control plane, the main issue is to solve the Multicast
Routing and Wavelength Assignment (MC-RWA) problem. The MC-RWA problem
involves establishing the multicast routes on the network, and determining the appro-
priate wavelength to be assigned to them, minimising the resources required (usually,
the number of wavelengths). The combined problem is NP-complete, as proved by
Ali and Deogun in [9]. For this reason, the RWA problem is often decoupled into two
separate sub-problems: the routing sub-problem and the wavelength assignment sub-
problem. The routing sub-problem is still NP-complete as it involves the construction
FHM From elextronixs to optixsO envwling texhniques
of a Steiner Minimum Tree, but wavelength assignment can be solved in polynomial
time, as shown in [47] and [134]. RWA problems are usually formulated as a Mixed
Integer Linear Programming (MILP) problem [125, 183] and solved using optimisers,
such as CPLEX [107]. It is only possible to solve MILP problems to optimality for
very small networks. For realistic-sized networks, scalable heuristics as those proposed
in [124, 125, 183, 185] are always necessary.
For the interested reader references [95, 173, 221] present detailed surveys of the
literature on optical multicast.
VCG irvffix grooming
Most applications have bandwidth requirements that are far less than that provided by
a single optical wavelength (or lightpath). It is therefore economical to use a lightpath
to concurrently support multiple connections. The process of allocating sub-wavelength
traffic demands to specific lightpaths such that the resources are shared is known as
the traffic grooming problem [70, 106]. Traffic grooming refers to techniques used
to aggregate low-speed traffic streams onto high-speed wavelengths. As explained in
Chapter 3, these lightpaths can then be all-optically switched in the intermediate nodes
(optical bypass), and thus save energy. Wang et al. [206], for instance, show that traffic
grooming mechanisms together with optical bypass are a feasible solution to reduce
electrical ports consumption in IP routers.
There has been some work on the design and operation of optical networks to sup-
port traffic grooming of multicast applications. Most work emphasises the reduction
of the required number of wavelength channels, as [120], but there is also investiga-
tion on minimising the number of higher layer electronic equipment (for example, IP
ports) [198]. In the context of IPTV, most grooming is usually done at the source, with
the TV head-end injecting all TV channels into one set of wavelengths and delivering
it to the IP network core. Grooming can also be relevant to add local TV channels at
the network edge, inserting local channels to be distributed to specific regions only.
VCH Vggregvtey multixvst
The capacity of a light-tree in core optical networks is much higher than the bandwidth
required by most multicast flows. Therefore, it is not efficient to directly map a single
multicast flow into one light-tree. To increase network efficiency, Zhu et al. [223, 224]
have studied the problem of aggregating multicast flows in IP over optical networks.
VCH Vggregvtey multixvst FHN
In [223] the authors show the problem is NP-complete, propose an optimal Integer
Linear Programming (ILP) solution and an efficient heuristic. In [224] Zhu and Jue
extend this work by separating pay-per-view and other secure channels from the rest.
The optical multicast flow aggregation problem is related to the aggregated mul-
ticast problem in IP networks, as studied by Cui et al. [49, 59]. However, while the
objective of the former is to increase network efficiency, the motivation of the latter
is scalability. The key idea of aggregated multicast is to force multiple IP multicast
sessions to share a single distribution tree to reduce multicast state in routers.

geferenxes
[1] A. Abramson. ihz Hi“tory of izlzvi“ionA FMME to FNIF. McFarland & Company,
1987. 1
[2] A. Adams, J. Nicholas, and W. Siadak. Protocol Independent Multicast - Dense
Mode (PIM-DM): Protocol specification (revised). In gFX HNLH, 2005. 17
[3] Y. Agarwal, S. Hodges, R. Chandra, J. Scott, P. Bahl, and R. Gupta. Somniloquy:
Augmenting network interfaces to reduce PC energy usage. In chDI, Boston, MA,
Apr. 2009. 40
[4] C. Aggarwal, J. Wolf, and P. S. Yu. On optimal piggyback merging policies for
video-on-demand systems. In hIGbEigIXh, Philadelphia, PA, May 1996. 29
[5] K. Ahmad and A. Begen. IPTV and video networks in the 2015 timeframe: The
evolution to medianets. IEEE Xommunixvtion“ bvgvzinz, 47(12):68–74, 2009. 5,
22
[6] Alcatel. Alcatel-Lucent WaveStar OLS 1.6T product specification. hggc-//gialhel.
Vbm/AlVTgXlWTiXfgTe. [Online; accessed 16-04-2012]. 119
[7] S. Aleksic. Analysis of power consumption in future high-capacity network nodes.
IEEEDdhV Journvl of dptixvl Xommunixvtion“ vny cztfiorking, 1(3):245–258,
2009. 46, 121
[8] M. Ali. Optimization of splitting node placement in wavelength-routed optical
networks. IEEE Journvl on hzlzxtzy Vrzv“ in Xommunixvtion“, 20(8):1571–1579,
2002. 137
[9] M. Ali and J. S. Deogun. Cost-effective implementation of multicasting in
wavelength-routed networks. Journvl of aightfivvz izxhnology, 18(12):1628–1638,
2000. 136, 137
FIG gZFZgZcXZh
[10] AMD. White paper: Magic packet technology, 1998. 39
[11] G. Anastasi, I. Giannetti, and A. Passarella. A Bittorrent proxy for green Internet
file sharing: Design and experimental evaluation. Xomputzr Xommunixvtion“,
33(7):794–802, 2010. 40
[12] C. Anderson. ihz aong ivilO lhy thz Futurz of Wu“inz““ i“ hzlling az““ of borz.
Hyperion, 2006. 83
[13] M. Andrews, A. F. Anta, L. Zhang, and W. Zhao. Routing and scheduling for
energy and delay minimization in the powerdown model. In IcFdXdb, San
Diego, CA, Mar. 2010. 42
[14] J. Baliga, R. Ayre, K. Hinton, W. V. Sorin, and R. S. Tucker. Energy consumption
in optical IP networks. Journvl of aightfivvz izxhnology, 27(13):2391–2403, 2009.
20, 37, 82, 107, 110, 118, 121
[15] J. Baliga, R. Ayre, K. Hinton, and R. S. Tucker. Photonic switching and the
energy bottleneck. In eh, San Francisco, CA, Aug. 2007. 37, 45, 110
[16] J. Baliga, R. Ayre, K. Hinton, and R. S. Tucker. Architectures for energy-efficient
IPTV networks. In dFX, San Diego, CA, Mar. 2009. 41
[17] J. Baliga, R. Ayre, W. V. Sorin, K. Hinton, and R. S. Tucker. Energy consumption
in access networks. In dFX, San Diego, CA, Feb. 2008. 37, 40
[18] J. Baliga, R. W. A. Ayre, K. Hinton, and R. S. Tucker. Green cloud computing:
Balancing energy in processing, storage, and transport. eroxzzying“ of thz IEEE,
99(1):149–167, January 2011. 41
[19] J. Baliga, K. Hinton, and R. S. Tucker. Energy consumption of the Internet. In
XdIcBVXdFi, Melbourne, Australia, June 2007. 37
[20] A. Ballardie. Core Based Trees (CBT version 2) multicast routing. In gFX GFMN,
1997. 18
[21] T. Ballardie, P. Francis, and J. Crowcroft. Core based trees (CBT). In hIGB
Xdbb, San Francisco, CA, Sept. 1993. 18, 131
[22] A. Banerjee, L. Drake, L. Lang, B. Turner, D. Awduche, L. Berger, K. Kompella,
and Y. Rekhter. Generalized multiprotocol label switching: an overview of sig-
naling enhancements and recovery techniques. IEEE Xommunixvtion“ bvgvzinz,
39(7):144–151, 2001. 47
gZFZgZcXZh FIH
[23] D. Banodkar, K. K. Ramakrishnan, S. Kalyanaraman, A. Gerber, and
O. Spatscheck. Multicast instant channel change in IPTV systems. In XdbB
hlVgE, Bangalore, India, Jan. 2008. 32
[24] L. A. Barroso and U. Holzle. The case for energy-proportional computing. IEEE
Xomputzr, 40(12):33–37, 2007. 38, 85, 93
[25] BBC. BBC iPlayer. hggc-//jjj.bbV.Vb.hk/iclTlXe/. [Online; accessed 16-04-2012].
5
[26] BBC. Niche TV growing quickly online. hggc-//aXjf.bbV.Vb.hk/2/hi/cebgeTmmXf/
VliVk baliaX/5108980.fgm. [Online; accessed 16-04-2012]. 82
[27] A. C. Begen, N. Glazebrook, and W. V. Steeg. Reducing channel-change times in
IPTV with Real-time Transport Protocol. IEEE Intzrnzt Xomputing, 13/3(3):40–
47, 2009. 12, 30
[28] A. C. Begen, N. Glazebrook, and W. V. Steeg. A unified approach for repairing
packet loss and accelerating channel changes in multicast IPTV. In XXcX, Las
Vegas, NV, Jan. 2009. 11, 12, 13, 30
[29] A. C. Begen, C. Perkins, and J. Ott. On the use of RTP for monitoring and fault
isolation in IPTV. IEEE cztfiork, 24(2):14–19, 2010. 2, 19, 22
[30] Y. Bejerano and P. Koppol. Improving zap response time for IPTV. In IcFdB
Xdb, Rio de Janeiro, Brazil, Apr. 2009. 14, 32
[31] A. P. Bianzino, C. Chaudet, D. Rossi, and J.-L. Rougier. A survey of green
networking research. Xogg, abs/1010.3880, 2010. 44
[32] A. P. Bianzino, A. K. Raju, and D. Rossi. Apples-to-apples: a framework analysis
for energy-efficiency in networks. hIGbEigIXh ezrformvnxz Evvluvtion gzvizfi,
38:81–85, January 2011. 43
[33] N. Bila, E. de Lara, M. Hiltunen, K. Joshi, H. A. Lagar-Cavilla, and M. Satya-
narayanan. The case for energy-oriented partial desktop migration. In HotXlouy
fiork“hop, Boston, MA, June 2010. 36, 40
[34] J. Blackburn and K. Christensen. A simulation study of a new green BitTorrent.
In IXX, Desden, Germany, June 2009. 42
FII gZFZgZcXZh
[35] J. Boyce and A. M. Tourapis. Fast efficient channel change. In IXXE, Singapore,
Nov. 2005. 33
[36] C.-S. Bres, N. Alic, E. Myslivets, and S. Radic. Scalable multicasting in one-pump
parametric amplifier. Journvl of aightfivvz izxhnology, 27(3):356–363, 2009. 136
[37] E. V. Breusegern, J. Cheyns, D. D. Winter, D. Colle, M. Pickavet, F. D. Turck,
and P. Demeester. Overspill routing in optical networks: a true hybrid optical
network design. IEEE Journvl on hzlzxtzy Vrzv“ in Xommunixvtion“, 24(4):13–
25, 2006. 47
[38] T. W. Cable. TWCable TV for iPad. hggc-//jjj.gimXjTeaXeVTblX.Vbm/alaj/
lXTea/Tccf/gjVgi/. [Online; accessed 16-04-2012]. 2
[39] B. Cain, S. Deering, I. Kouvelas, B. Fenner, and A. Thyagarajan. Internet Group
Management Protocol, version 3. In gFX HHLK, 2002. 11, 16
[40] J. Caja. Optimization of IPTV multicast traffic transport over next generation
metro networks. In cEildgKh, New Delhi, India, June 2006. 83
[41] Y. Cao and O. Yu. Optimal placement of light splitters and wavelength converters
for multicast in WDM networks. In IXXVh, Busan, Korea, Sept. 2005. 137
[42] M. Cha, P. Rodriguez, J. Crowcroft, S. Moon, and X. Amatriain. Watching
television over an IP network. In IbX, Vouliagmeni, Greece, Oct. 2008. 1, 19,
28, 50, 57, 69, 71, 72, 76, 83, 102, 110
[43] M. Cha, P. Rodriguez, S. Moon, and J. Crowcroft. On next-generation telco-
managed P2P TV architectures. In Ieieh, Tampa Bay, FL, Feb. 2008. 3, 28
[44] J. Chabarek, J. Sommers, P. Barford, C. Estan, D. Tsiang, and S. Wright. Power
awareness in network design and routing. In IcFdXdb, Phoenix, AZ, Apr. 2008.
43, 93, 95
[45] P. Chanclou, S. Gosselin, J. F. Palacios, V. L. Alvarez, and E. Zouganeli.
Overview of the optical broadband access evolution: a joint article by opera-
tors in the IST network of excellence e-Photon/One. IEEE Xommunixvtion“
bvgvzinz, 44(8):29–35, 2006. 21
[46] C. S. Chang, D. S. Lee, and Y. S. Jou. Load balanced Birkhoff-von Neumann
switches. In lork“hop on High ezrformvnxz hfiitxhing vny gouting, Dallas, TX,
May 2001. 44
gZFZgZcXZh FI5
[47] B. Chen and J. Wang. Efficient routing and wavelength assignment for multi-
cast in WDM networks. IEEE Journvl on hzlzxtzy Vrzv“ in Xommunixvtion“,
20(1):97–109, 2002. 138
[48] S. Chen, T. H. Cheng, and G.-S. Poo. Placement of wavelength converters and
light splitters in a WDM network using the generic graph model. Xomputzr
Xommunixvtion, 33(7):868–883, 2010. 137
[49] J.-H. Chi, M. Gerla, K. Boussetta, M. Faloutsos, A. Fei, J. Kim, and D. Mag-
giorini. Aggregated multicast: A scheme to reduce multicast states. Technical
report, Internet Engineering Task Force, 2002. 139
[50] L. Chiaraviglio, M. Mellia, and F. Neri. Energy-aware backbone networks: A
case study. In IXX, Desden, Germany, June 2009. 38
[51] C. Cho, I. Han, Y. Jun, and H. Lee. Improvement of channel zapping time in
IPTV services using the adjacent groups join-leave method. In IXVXi, Gangwon-
Do, Korea, Feb. 2004. 34, 64
[52] K. J. Christensen, C. Gunaratne, B. Nordman, and A. D. George. The next
frontier for communications networks: power management. Xomputzr XommuB
nixvtion“, 27(18):1758–1770, 2004. 39
[53] M. Christensen, K. Kimball, and F. Solensky. Considerations for Internet Group
Management Protocol (IGMP) and Multicast Listener Discovery (MLD) snooping
switches. In gFX IJIF, 2006. 16
[54] A. Cianfrani, V. Eramo, M. Listanti, M. Marazza, and E. Vittorini. An energy
saving routing algorithm for a green OSPF protocol. In IcFdXdb, San Diego,
CA, Mar. 2010. 42
[55] J. M. Cioffi, H. Zou, A. Chowdhery, W. Lee, and S. Jagannathan. Greener copper
with dynamic spectrum management. In GadWEXdb, New Orleans, LA, Nov.
2008. 40
[56] Cisco. Cisco visual networking index: Forecast and methodology 2008-2013. 3
[57] E. Commission. Growth of the number of TV channels and multi-channel plat-
forms in Europe continues despite the crisis. 82
[58] J. Crowcroft, M. Handley, and I. Wakeman. Intzrnztfiorking bultimzyiv. Morgan
Kaufmann, 1999. 15
FI6 gZFZgZcXZh
[59] J.-H. Cui, J. Kim, D. Maggiorini, K. Boussetta, and M. Gerla. Aggregated
multicast - a comparative study. Xlu“tzr Xomputing, 8(1):15–26, January 2005.
139
[60] C. da Manha (Portuguese newspaper). Meo ultrapassa os dez mil canais (trvn“lvB
tionO bzo vlrzvyy hv“ ovzr tzn thou“vny xhvnnzl“). hggc-//gialhel.Vbm/MXbGXaK.
[Online; accessed 16-04-2012]. 83
[61] T. Das, P. Padala, V. N. Padmanabhanand, R. Ramjee, and K. G. Shin. Lite-
Green: saving energy in networked desktops using virtualization. In jhEcIm
ViX, Boston, MA, June 2010. 36, 40
[62] S. Deering. Host extensions for IP multicasting. In gFX FFFG, 1989. 14, 15, 135
[63] S. Deering, D. L. Estrin, D. Farinacci, V. Jacobson, C.-G. Liu, and L. Wei. The
PIM architecture for wide-area multicast routing. IEEEDVXb irvn“vxtion“ on
cztfiorking, 4(2):153–162, 1996. 130
[64] S. E. Deering. bultixv“t gouting in v Dvtvgrvm Intzrnztfiork. PhD thesis, Stan-
ford University, 1991. 15
[65] N. Degrande, K. Laevens, D. D. Vleeschauwer, and R. Sharpe. Increasing the user
perceived quality for IPTV services. IEEE Xommunixvtion“ bvgvzinz, 46(2):94–
100, 2008. 9, 11, 33
[66] D.-R. Din and C.-Y. Li. A genetic algorithm for solving virtual source placement
problem on WDM networks. Xomputzr Xommunixvtion, 32(2):397–408, 2009.
137
[67] C. Diot, B. N. Levine, B. Lyles, H. Kassem, and D. Balensiefen. Deployment
issues for the IP multicast service and architecture. IEEE cztfiork, 14(1):78–88,
2000. 23
[68] X. Dong, T. El-Gorashi, and J. M. H. Elmirghani. IP over WDM networks
employing renewable energy sources. Journvl of aightfivvz izxhnology, 29(1):3,
2011. 36, 43, 132
[69] R. Doverspike, K. Ramakrishnan, and C. Chase. Structural overview of ISP
networks. In Guiyz to gzlivwlz Intzrnzt hzrvixz“ vny Vpplixvtion“. Springer, 2010.
95
gZFZgZcXZh FIL
[70] R. Dutta and G. N. Rouskas. Traffic grooming in WDM networks: past and
future. IEEE cztfiork, 16(6):46–56, 2002. 138
[71] Economist. In praise of television: the great survivor. hggc-//jjj.XVbabmifg.
Vbm/abWX/16009155. [Online; accessed 16-04-2012]. 3
[72] A. Feldmann, A. Gladisch, M. Kind, C. Lange, G. Smaragdakis, and F.-J. West-
phal. Energy trade-offs among content delivery architectures. In XiiE, Ghent,
Belgium, June 2010. 41
[73] B. Fenner, M. Handley, H. Holbrook, and I. Kovelas. Protocol Independent
Multicast - Sparse Mode (PIM-SM): Protocol specification (revised). In gFX
IKEF, 2006. 18, 21, 130
[74] W. Fenner. Internet Group Management Protocol, version 2. In gFX GGHK, 1997.
16
[75] G. M. Fernandez, C. Vazquez, P. C. Lallana, and D. Larrabeiti. Tap-and-2-split
switch design based on integrated optics for light-tree routing in WDM networks.
Journvl of aightfivvz izxhnology, 27(13):2506–2517, 2009. 136
[76] M. Fiorani, M. Casoni, and S. Aleksic. Performance and power consumption
analysis of a hybrid optical core node. IEEEDdhV Journvl of dptixvl XommuniB
xvtion“ vny cztfiorking, 3(6):502–513, 2011. 121
[77] W. Fisher, M. Suchara, and J. Rexford. Greening backbone networks: Reduc-
ing energy consumption by shutting off cables in bundled links. In hIGXdbb
fiork“hop on grzzn nztfiorking, New Delhi, India, Aug. 2010. 38
[78] M. P. Fok and C. Shu. Performance investigation of one-to-six wavelength mul-
ticasting of ASK-DPSK signal in a highly nonlinear bismuth oxide fiber. Journvl
of aightfivvz izxhnology, 27(15):2953–2957, 2009. 136
[79] C. Fraleigh, S. Moon, B. Lyles, C. Cotton, M. Khan, D. Moll, R. Rockell, T. Seely,
and S. C. Diot. Packet-level traffic measurements from the Sprint IP backbone.
IEEE cztfiork, 17(6):6–16, 2003. 19, 20, 87
[80] H. Fuchs and N. Farber. Optimizing channel change time in IPTV applications.
In WbhW, Las Vegas, NV, Apr. 2008. 12, 13, 62
[81] C.-H. Gan, P. Lin, and C.-M. Chen. A novel prebuffering scheme for IPTV
service. Xomputzr cztfiork“, 53(11):1956–1966, 2009. 35, 49, 61
FIM gZFZgZcXZh
[82] C. M. Gauger, P. J. Kuhn, E. V. Breusegem, M. Pickavet, and P. Demeester.
Hybrid optical network architectures: bringing packets and circuits together.
IEEE Xommunixvtion“ bvgvzinz, 44(8):36–42, 2006. 47
[83] M. Ghanbari, D. Crawford, M. Fleury, E. Khan, J. Woods, H. Lu, and R. Ravazi.
Future performance of video codecs. Technical report, University of Essex, 2006.
9, 23, 82, 91
[84] M. Ghandour, S. Liu, D. K. Hunter, D. Simeonidou, R. Nejabati, and P. Petropou-
los. Ultra high performance media multicasting scheme over wavelength-routed
networks. In dFX, San Diego, CA, Mar. 2010. 136
[85] N. Ghani, S. Dixit, and T.-S. Wang. On IP-over-WDM integration. IEEE XomB
munixvtion“ bvgvzinz, 38(3):72–84, March 2000. 109
[86] G. Gilder. aifz vftzr izlzvi“ion. W. W. Norton & Company, 1992. 3
[87] F. Glover. Tabu search — a tutorial. Intzrfvxz“, 20(4):74–94, 1990. 131
[88] A. Greenberg, G. Hjalmtysson, and J. Yates. Smart routers-simple optics a net-
work architecture for IP over WDM. In dFX, Baltimore, MD, Mar. 2000. 47
[89] GreenTouch. Greentouch website. hggc-//jjj.geXXagbhVh.beg/. [Online; accessed
16-04-2012]. 82
[90] C. Gunaratne, K. Christensen, and B. Nordman. Managing energy consump-
tion costs in desktop PCs and LAN switches with proxying, split TCP connec-
tions, and scaling of link speed. Intzrnvtionvl Journvl of cztfiork bvnvgzmznt,
15(5):297–310, 2005. 39
[91] C. Gunaratne, K. Christensen, B. Nordman, and S. Suen. Reducing the energy
consumption of Ethernet with adaptive link rate (ALR). IEEE irvn“vxtion“ on
Xomputzr“, 57(4):448–461, 2008. 39
[92] M. Gupta, S. Grover, and S. Singh. A feasibility study for power management in
LAN switches. In IXce, Berlin, Germany, Oct. 2004. 38
[93] M. Gupta and S. Singh. Greening of the Internet. In hIGXdbb, Karlsruhe,
Germany, Aug. 2003. 35, 36, 38
[94] M. Gupta and S. Singh. Using low-power modes for energy conservation in Eth-
ernet LANs. In IcFdXdb, Anchorage, AK, May 2007. 38
gZFZgZcXZh FIN
[95] A. Hamad, T. Wu, A. E. Kamal, and A. K. Somani. On multicasting in
wavelength-routing mesh networks. Xomputzr cztfiork“, 50(16):3105–3164, 2006.
138
[96] A. M. Hamad and A. E. Kamal. Optimal power-aware design of all-optical mul-
ticasting in wavelength routed networks. In IXX, Paris, France, June 2004. 136
[97] A. M. Hamad and A. E. Kamal. Optical amplifiers placement in WDM mesh
networks for optical multicasting service support. IEEEDdhV Journvl of dptixvl
Xommunixvtion“ vny cztfiorking, 1(1):85–102, 2009. 136
[98] B. Heller, S. Seetharaman, P. Mahadevan, Y. Yiakoumis, P. Sharma, S. Banerjee,
and N. McKeown. ElasticTree: saving energy in data center networks. In chDI,
San Jose, CA, Apr. 2010. 40
[99] K. Hinton, J. Baliga, R. Ayre, and R. Tucker. The future Internet — an energy
consumption perspective. In dEXX, Hong Kong, China, July 2009. 45
[100] K. Hinton, G. Raskutti, P. M. Farrell, and R. S. Tucker. Switching energy and
device size limits on digital photonic signal processing technologies. IEEE Journvl
of hzlzxtzy iopix“ in fuvntum Elzxtronix“, 14(3):938–945, 2008. 45
[101] G. Hjalmtysson, J. Yates, S. Chaudhuri, and A. Greenberg. Smart routers-simple
optics: an architecture for the optical Internet. Journvl of aightfivvz izxhnology,
18(12):1880–1891, 2000. 47
[102] H. Hlavacs and S. Buchinger. Hierarchical video patching with optimal server
bandwidth. VXb irvn“vxtion“ on bultimzyiv XomputingA Xommunixvtion“A vny
Vpplixvtion“, 4(1):1–23, 2008. 30
[103] H. Holbrook, B. Cain, and B. Haberman. Using Internet Group Management
Protocol version 3 (IGMPv3) and Multicast Listener Discovery Protocol version
2 (MLDv2) for Source-Specific Multicast. In gFX IKEI, 2006. 16
[104] W. Hou, L. Guo, J. Cao, J. Wu, and L. Hao. Green multicast grooming based on
optical bypass technology. dptixvl Fiwzr izxhnology, 17(2):111–119, 2011. 110,
119
[105] H. Huang and J. A. Copeland. Optical networks with hybrid routing. IEEE
Journvl on hzlzxtzy Vrzv“ in Xommunixvtion“, 21(7):1063–1070, 2003. 47
F5E gZFZgZcXZh
[106] S. Huang and R. Dutta. Dynamic traffic grooming: the changing role of traffic
grooming. IEEE Xommunixvtion“ hurvzy“ vny iutorivl“, 9(1):32–50, 2007. 138
[107] IBM. ILOG CPLEX optimizer. hggc-//gialhel.Vbm/I5M-6CL8K. [Online; accessed
16-04-2012]. 138
[108] F. Idzikowski, S. Orlowski, C. Raack, H. Woesner, and A. Wolisz. Saving energy
in IP-over-WDM networks by switching off line cards in low-demand scenarios.
In dcDb, Copenhagen, Denmark, May 2010. 38
[109] ISO/IEC. Generic coding of moving pictures and associated audio information.
In IhdDIEX FHMFM, 1995. 9
[110] ISO/IEC. Coding of audio-visual objects. In IhdDIEX FIINK, 1998. 9
[111] S. Iyer, S. Bhattacharyya, N. Taft, and C. Diot. An approach to alleviate link
overload as observed on an IP backbone. In IcFdXdb, San Franciso, CA, Mar.
2003. 97
[112] V. Jacobson, D. Smetters, J. Thornton, M. Plass, N. Briggs, and R. Braynard.
Networking named content. In XocEmi, Rome, Italy, Dec. 2009. 41
[113] R. Jain. ihz vrt of xomputzr “y“tzm“ pzrformvnxz vnvly“i“O izxhniquz“ for zflB
pzrimzntvl yz“ignA mzv“urzmzntA “imulvtionA vny moyzling. Wiley-Interscience,
1991. 49, 56, 60
[114] U. Jennehag and S. Pettersson. On synchronization frames for channel switching
in a GOP-based IPTV environment. In XXcX, Las Vegas, NV, Jan. 2008. 34
[115] U. Jennehag, T. Zhang, and S. Pettersson. Improving transmission efficiency in
H.264 based IPTV systems. IEEE irvn“vxtion“ on Wrovyxv“ting, 53(1):69–78,
2007. 33
[116] M. Jimeno and K. Christensen. A prototype power management proxy for
Gnutella peer-to-peer file sharing. In aXc, Dublin, Ireland, Oct. 2007. 40
[117] H. Joo, H. Song, D.-B. Lee, and I. Lee. An effective IPTV channel control algo-
rithm considering channel zapping time and network utilization. IEEE irvn“vxB
tion“ on Wrovyxv“ting, 54(2):208–216, 2009. 34
[118] M. Kalman, S. T. Eckehard, and B. Girod. Adaptive playout for real-time media
streaming. In IhXVh, Scottsdale, AZ, May 2002. 29, 130
gZFZgZcXZh F5F
[119] M. Kalman, E. Steinbach, and B. Girod. Adaptive media playout for low delay
video streaming over error-prone channels. IEEE irvn“vxtion“ on Xirxuit“ vny
hy“tzm“ for kiyzo izxhnology, 14(6):841–851, 2004. 29, 130
[120] A. E. Kamal. Algorithms for multicast traffic grooming in WDM mesh networks.
IEEE Xommunixvtion“ bvgvzinz, 44(11):96–105, 2006. 138
[121] M. Karczewicz and R. Kurceren. The SP- and SI-frames design for H.264/AVC.
IEEE irvn“vxtion“ on Xirxuit“ vny hy“tzm“ for kiyzo izxhnology, 13(7):637–644,
2003. 33
[122] R. H. Katz. Tech titans building boom. IEEE hpzxtrum, 46(2):40–54, 2009. 36
[123] I. Keslassy, S.-T. Chuang, K. Yu, D. Miller, M. Horowitz, O. Solgaard, and
N. McKeown. Scaling internet routers using optics. In hIGXdbb, Karlsruhe,
Germany, Aug. 2003. 44
[124] F. Koksal and C. Ersoy. A flexible scalable solution for all-optical multifiber
multicasting: SLAM. Journvl of aightfivvz izxhnology, 25(9):2653–2666, 2007.
138
[125] F. Ko¨ksal and C. Ersoy. Multicasting for all-optical multifiber networks. Journvl
of dptixvl cztfiorking, 6(2):219–238, 2007. 138
[126] K. Kompella and Y. Rekhter. OSPF extensions in support of Generalized Multi-
Protocol Label Switching (GMPLS). In gFX IGEH, 2005. 47
[127] K. Kompella and Y. Rekhter. Routing extensions in support of Generalized
Multi-Protocol Label Switching (GMPLS). In gFX IGEG, 2005. 47
[128] R. Kooij, K. Ahmed, K. Brunnstro¨m, and K. Acreo. Perceived quality of channel
zapping. In Xhc, Palma de Mallorca, Spain, Aug. 2006. 11, 63
[129] I. Kopilovic and M. Wagner. A benchmark for fast channel change in IPTV. In
WbhW, Las Vegas, NV, Apr. 2008. 29
[130] C. Y. Lee, C. K. Hong, and K. Y. Lee. Reducing channel zapping time in IPTV
based on user’s channel selection behaviors. IEEE irvn“vxtion“ on Wrovyxv“ting,
56(3):321–330, 2010. 35, 49
F5G gZFZgZcXZh
[131] E. Lee, J. Whang, U. Oh, K. Koh, and H. Bahn. Popular channel concentration
schemes for efficient channel navigation in Internet Protocol televisions. IEEE
irvn“vxtion“ on Xon“umzr Elzxtronix“, 55(4):1945–1949, 2009. 35
[132] U. Lee, I. Rimac, and V. Hilt. Greening the Internet with content-centric net-
working. In zBEnzrgy, Passau, Germany, Apr. 2010. 41
[133] F. Legendre, V. Lenders, M. May, and G. Karlsson. Narrowcasting: an empirical
performance evaluation study. In XHVcih, San Francisco, CA, Sept. 2008. 82
[134] R. Libeskind-Hadas and R. Melhem. Multicast routing and wavelength assign-
ment in multihop optical networks. IEEEDVXb irvn“vxtion“ on cztfiorking,
10(5):621–629, 2002. 138
[135] H.-C. Lin and S.-W. Wang. Splitter placement in all-optical WDM networks. In
GadWEXdb, St. Louis, MO, Nov. 2005. 137
[136] P. Mahadevan, P. Sharma, S. Banerjee, and P. Ranganathan. A power bench-
marking framework for network devices. In IFIe, Buenos Aires, Argentina, June
2009. 43, 92
[137] A. Mahimkar, Z. Ge, A. Shaikh, J. Wang, J. Yates, Y. Zhang, and Q. Zhao. To-
wards automated performance diagnosis in a large IPTV network. In hIGXdbb,
Barcelona, Spain, Aug. 2009. 29
[138] J. Maisonneuve, M. Deschanel, J. Heiles, W. Li, H. Liu, R. Sharpe, and Y. Wu. An
overview of IPTV standards development. IEEE irvn“vxtion“ on Wrovyxv“ting,
55(2):315–328, 2009. 3
[139] G. Malkin. RIP version 2. In gFX GIJH, 1998. 17
[140] F. Mannie. Generalized Multi-Protocol Label Switching (GMPLS) architecture.
In gFX HNIJ, 2004. 47
[141] Microsoft. Delivering IPTV with the Windows Media Platform. Technical report,
Microsoft, 2003. 2, 30, 129
[142] P. V. Mieghem. ezrformvnxz Vnvly“i“ of Xommunixvtion“ cztfiork“ vny hy“tzm“.
Cambridge University Press, 2009. 110
gZFZgZcXZh F5H
[143] Y. Morita, H. Hasegawa, K.-I. S. Y., Sone, K. Yamada, and M. Jinno. Optical
multicast tree construction algorithm considering SNR constraint and 3R regen-
eration. In EXdX, Brussels, Belgium, Sept. 2008. 137
[144] S. Nedevschi, J. Chandrashekar, J. Liu, B. Nordman, S. Ratnasamy, and N. Taft.
Skilled in the art of being idle: reducing energy waste in networked systems. In
chDI, Boston, MA, Apr. 2009. 39
[145] S. Nedevschi, L. Popa, G. Iannaccone, S. Ratnasamy, and D. Wetherall. Reducing
network energy consumption via sleeping and rate-adaptation. In chDI, San
Francisco, CA, Apr. 2008. 38
[146] NetFPGA. Netfpga website. hggc-//aXgfcgT.beg/. [Online; accessed 16-04-2012].
43
[147] Nielsen. The Nielsen company. hggc-//jjj.aiXlfXa.Vbm/. [Online; accessed 16-
04-2012]. 28, 71
[148] Nielsen. Nielsen’s Q1 2010 three screen report. Technical report, The Nielsen
Company, 2006. 3
[149] A. Odlyzko. Data networks are lightly utilized, and will stay that way. gzvizfi
of cztfiork Exonomix“, 2(3), 2003. 39
[150] U. Oh, S. Lim, and H. Bahn. Channel reordering and prefetching schemes for
efficient IPTV channel navigation. IEEE irvn“vxtion“ on Xon“umzr Elzxtronix“,
56(2):483–487, 2010. 35, 49, 73
[151] J. Ott, S. Wenger, N. Sato, C. Burmeister, and J. Rey. Extended RTP profile for
Real-time Transport Control Protocol (RTCP)-based feedback (RTP/AVPF). In
gFX IJMJ, 2006. 30
[152] F. Palmieri. GMPLS control plane services in the next-generation optical internet.
ihz Intzrnzt erotoxol Journvl, 11(3):2–18, 2008. 47
[153] C. Panarello, A. Lombardo, G. Schembra, L. Chiaraviglio, and M. Mellia. Energy
saving and network performance: a trade-off approach. In zBEnzrgy, Passau,
Germany, Apr. 2010. 40
[154] D. Papadimitriou and A. Farrel. Generalized Multi-Protocol Label Switching
(GMPLS) Signaling Resource ReserVation Protocol-Traffic Engineering (RSVP-
TE) extensions. In gFX HILH, 2003. 47, 112
F5I gZFZgZcXZh
[155] M. Pervila¨ and J. Kangasharju. Running servers around zero degrees. In hIGB
Xdbb fiork“hop on grzzn nztfiorking, New Delhi, India, Aug. 2010. 36
[156] B. Piper. United states IPTV market sizing: 2009-2013. Technical report, Strat-
egy Analytics, 2009. 3
[157] O. B. Portal. Average advertised download speeds, by country. 65, 78
[158] J. Postel. User Datagram Protocol. In gFX LKM, 1980. 13
[159] PT. Meo customers may now have their own TV channel. hggc-//gialhel.Vbm/
MXbBjaGI. [Online; accessed 16-04-2012]. 82
[160] T. Qiu, Z. Ge, S. Lee, J. Wang, J. Xu, and Q. Zhao. Modeling user activities in
a large IPTV system. In IbX, Chicago, IL, Nov. 2009. 28, 35, 49, 50, 56, 59, 64,
76, 82, 91, 120
[161] T. Qiu, Z. Ge, S. Lee, J. Wang, Q. Zhao, and J. Xu. Modeling channel popularity
dynamics in a large IPTV system. In hIGbEigIXh, Seattle, WA, June 2009.
28, 56, 57, 59, 72, 76, 83, 102, 107, 110, 115
[162] R. Ramaswami, K. Sivarajan, and G. Sasaki. dptixvl cztfiork“O V ervxtixvl
ezr“pzxtivz. Morgan Kaufmann, 2009. 106, 107, 137
[163] F. M. V. Ramos, J. Crowcroft, R. J. Gibbens, P. Rodriguez, and I. H. White.
Channel smurfing: minimising channel switching delay in IPTV distribution net-
works. In IXbE, Singapore, July 2010. 6
[164] F. M. V. Ramos, J. Crowcroft, R. J. Gibbens, P. Rodriguez, and I. H. White.
Reducing channel change delay in IPTV by predictive pre-joining of TV channels.
hignvl eroxz““ingO Imvgz Xommunixvtion, 26(7):400–412, 2011. 6
[165] F. M. V. Ramos, R. J. Gibbens, F. Song, P. Rodriguez, J. Crowcroft, and I. H.
White. Reducing energy consumption in IPTV networks by selective pre-joining
of channels. In hIGXdbb fiork“hop on grzzn nztfiorking, New Delhi, India,
Aug. 2010. 7
[166] F. M. V. Ramos, F. Song, P. Rodriguez, R. Gibbens, J. Crowcroft, and I. H.
White. Constructing an IPTV workload model. In hIGXdbb eo“tzr hz““ion,
Barcelona, Spain, Aug. 2009. 35, 72
gZFZgZcXZh F55
[167] P. Ranganathan. Recipe for efficiency: principles of power-aware computing.
Xommunixvtion“ of thz VXb, 53(4):60–67, 2010. 81, 103
[168] Y. Rekhter, T. Li, and S. Hares. A Border Gateway Protocol 4 (BGP-4). In gFX
IGLF, 2006. 5
[169] J. C. C. Restrepo, C. G. Gruber, and C. M. Machuca. Energy profile aware
routing. In IXX, Desden, Germany, June 2009. 42
[170] Reuters. Going niche is the future for Indian television. hggc-//ia.eXhgXef.Vbm/
TegiVlX/2010/10/20/iWIAIaWiT-52320320101020. [Online; accessed 16-04-2012]. 82
[171] J. Rey, D. Leon, A. Miyazaki, V. Varsa, and R. Hakenberg. RTP retransmission
payload format. In gFX IJMM, 2006. 30
[172] E. Rosen, A. Viswanathan, and R. Callon. Multiprotocol Label Switching Archi-
tecture. In gFX HEHF, 2001. 47
[173] G. N. Rouskas. Optical layer multicast: Rationale, building blocks, and chal-
lenges. IEEE cztfiork, 17(1):60–65, 2003. 138
[174] L. H. Sahasrabuddhe and B. Mukherjee. Light trees: optical multicasting for
improved performance in wavelength routed networks. IEEE Xommunixvtion“
bvgvzinz, 37(2):67–73, 1999. 135
[175] A. Saleh and J. Simmons. Evolution toward the next-generation core optical
network. Journvl of aightfivvz izxhnology, 24(9):3303–3321, 2006. 45, 106, 107
[176] C. Sasaki, A. Tagami, T. Hasegawa, and S. Ano. Rapid channel zapping for
IPTV broadcasting with additional multicast stream. In IXX, Beijing, China,
May 2008. 11, 32, 62
[177] K. Sato, N. Yamanaka, Y. Takigawa, M. Koga, S. Okamoto, K. Shiomoto, E. Oki,
and W. Imajuku. GMPLS-based photonic multilayer router (Hikari router) ar-
chitecture: an overview of traffic engineering and signaling technology. IEEE
Xommunixvtion“ bvgvzinz, 40(3):96–101, 2002. 47
[178] P. Savola. Overview of the Internet multicast routing architecture. In gFX JFFE,
2008. 18, 21, 130
[179] G. Shen and R. S. Tucker. Translucent optical networks: the way forward. IEEE
Xommunixvtion“ bvgvzinz, 45(2):48–54, 2007. 46
F56 gZFZgZcXZh
[180] G. Shen and R. S. Tucker. Energy-minimized design for IP over WDM networks.
IEEEDdhV Journvl of dptixvl Xommunixvtion“ vny cztfiorking, 1(1):176–186,
2009. 43, 45, 107, 109, 110, 119, 121
[181] S. W. Sherman and J. C. Browne. Trace driven modeling: Review and overview.
In Vchh, 1973. 56
[182] P. Siebert, T. N. M. V. Caenegem, and M. Wagner. Analysis and improvements of
zapping times in IPTV systems. IEEE irvn“vxtion“ on Wrovyxv“ting, 55(2):407–
418, June 2009. 10, 11, 12, 13, 62
[183] N. K. Singhal, L. H. Sahasrabuddhe, and B. Mukherjee. Optimal multicasting
of multiple light-trees of different bandwidth granularities in a WDM mesh net-
work with sparse splitting capabilities. IEEEDVXb irvn“vxtion“ on cztfiorking,
14(5):1104–1117, 2006. 138
[184] V. Sivaraman, A. Vishwanath, Z. Zhao, and C. Russell. Profiling per-packet and
per-byte energy consumption in the NetFPGA Gigabit router. In IcFdXdb
lork“hop on Grzzn Xommunixvtion“ vny cztfiorking, Shanghai, China, Apr.
2011. 43, 84, 92, 93, 95, 108, 121
[185] N. Skorin-Kapov. Multicast routing and wavelength assignment in WDM net-
works: a binpacking approach. Journvl of dptixvl cztfiorking, 5(4):266–279,
2006. 137, 138
[186] D. Smith. IPTV bandwidth demand: Multicast and channel surfing. In IcFdB
Xdb, Anchorage, AK, May 2007. 31
[187] E. Steinbach, N. Farber, and B. Girod. Adaptive playout for low latency video
streaming. In IXIe, Thessaloniki, Greece, Oct. 2001. 130
[188] I. Stoica, T. S. E. Ng, and H. Zhang. REUNITE: a recursive unicast approach
to multicast. In IcFdXdb, Tel-Aviv, Israel, Mar. 2000. 131
[189] W. Sun, K. Lin, and Y. Guan. Performance analysis of a finite duration
multichannel delivery method in IPTV. IEEE irvn“vxtion“ on Wrovyxv“ting,
54(3):419–429, 2008. 12, 13, 35, 49, 61, 62, 67
[190] A. S. Tanenbaum and M. V. Steen. Di“triwutzy hy“tzm“O erinxiplz“ vny
evrvyigm“. Pearson Prentice Hall, 2007. 52
gZFZgZcXZh F5L
[191] G. Thompson and Y.-F. R. Chen. IPTV: Reinventing television in the internet
age. IEEE Intzrnzt Xomputing, 13(3):11–14, 2009. 1, 2, 21
[192] P. Tsiaflakis, Y. Yi, M. Chiang, and M. Moonen. Green DSL: Energy-efficient
DSM. In IXX, Desden, Germany, June 2009. 40
[193] R. S. Tucker. The role of optics and electronics in high-capacity routers. Journvl
of aightfivvz izxhnology, 24(12):4655–4673, 2006. 44, 121
[194] R. S. Tucker. Green optical communications Part I: Energy limitations in trans-
port. IEEE Journvl of hzlzxtzy iopix“ in fuvntum Elzxtronix“, 17(2):245–260,
2011. 37
[195] R. S. Tucker. Green optical communications Part II: Energy limitations in net-
works. IEEE Journvl of hzlzxtzy iopix“ in fuvntum Elzxtronix“, 17(2):261–274,
2011. 36, 37
[196] R. S. Tucker, K. Hinton, and G. Raskutti. Energy consumption limits in high-
speed optical and electronic signal processing. Elzxtronix“ azttzr“, 43(17):906–
908, 2007. 45
[197] R. S. Tucker, R. Parthiban, J. Baliga, K. Hinton, R. W. Ayre, and W. V. Sorin.
Evolution of WDM optical IP networks: A cost and energy perspective. Journvl
of aightfivvz izxhnology, 27(3):243–252, 2009. 37, 82
[198] R. Ul-Mustafa and A. E. Kamal. Design and provisioning of WDM networks with
multicast traffic grooming. IEEE Journvl on hzlzxtzy Vrzv“ in Xommunixvtion“,
24(4):37–53, 2006. 138
[199] H. Uzunalioglu. Channel change delay in IPTV systems. In XXcX, Las Vegas,
NV, Jan. 2009. 11
[200] V. Valancius, N. Laoutaris, L. Massoulie´, C. Diot, and P. Rodriguez. Greening
the Internet with nano data centers. In XocEmi, Rome, Italy, Dec. 2009. 41
[201] S. Vanhastel and R. Hernandez. Enabling IPTV: What’s needed in the access
network. IEEE Xommunixvtion“ bvgvzinz, 46(8):90–95, 2008. 3, 21, 24
[202] N. Vasic´ and D. Kostic´. Energy-aware traffic engineering. In zBEnzrgy, Passau,
Germany, Apr. 2010. 42
F5M gZFZgZcXZh
[203] D. Waitzman, C. Partridge, and S. Deering. Distance Vector Multicast Routing
Protocol. In gFX FELJ, 1988. 17
[204] H. Wang, X. Meng, M. Zhang, and Y. Li. Tabu search algorithm for RP selection
in PIM-SM multicast routing. Xomputzr Xommunixvtion“, 33(1):35–42, January
2010. 131
[205] H. Wang, A. Wonfor, K. A. Williams, R. V. Penty, and I. H. White. Demonstra-
tion of a lossless monolithic 16x16 QW SOA switch. In EXdX, Vienna, Austria,
Sept. 2009. 107
[206] X. Wang, W. Hou, L. Guo, J. Cao, and D. Jiang. Energy saving and cost reduction
in multi-granularity green optical networks. Xomputzr cztfiork“, 55(3):676–688,
2011. 138
[207] Y. Wang and Y. Yang. Multicasting in a class of multicast-capable WDM net-
works. Journvl of aightfivvz izxhnology, 20(3):350–359, 2002. 135
[208] T. Wiegand, L. Noblet, and F. Rovati. Scalable video coding for IPTV services.
IEEE irvn“vxtion“ on Wrovyxv“ting, 55(2):527–538, June 2009. 130
[209] Wikipedia. Paul Gottlieb Nipkow. hggc-//Xa.jikicXWiT.beg/jiki/CThl :bggliXb
Aickbj. [Online; accessed 16-04-2012]. 1
[210] Wired. CES 2008: Panasonics enormous 150” plasma TV dwarfs all competi-
tors. hggc-//jjj.jieXW.Vbm/gTWgXglTb/2008/01/VXf-2008-kXlabg/. [Online; ac-
cessed 16-04-2012]. xvii, 22, 23
[211] Y. Xiao, X. Du, J. Zhang, F. Hu, and S. Guizani. Internet Protocol television
(IPTV): The killer application for the next-generation Internet. IEEE XommuB
nixvtion“ bvgvzinz, 45(11):126–134, 2007. 1
[212] Y. Xin and G. Rouskas. Multicast routing under optical layer constraints. In
IcFdXdb, Hong Kong, China, Mar. 2004. 137
[213] D. Xu, M. Chiang, and J. Rexford. Link-state routing with hop-by-hop forwarding
can achieve optimal traffic engineering. In IcFdXdb, Phoenix, AZ, Apr. 2008.
130
[214] D.-N. Yang andW. Liao. Design of light-tree based logical topologies for multicast
streams in wavelength routed optical networks. In IcFdXdb, San Franciso, CA,
Mar. 2003. 137
gZFZgZcXZh F5N
[215] M. Yano, F. Yamagishi, and T. Tsuda. Optical MEMS for photonic switching-
compact and stable optical crossconnect switches for simple, fast, and flexible
wavelength applications in recent photonic networks. IEEE Journvl of hzlzxtzy
iopix“ in fuvntum Elzxtronix“, 11(2):383–394, 2005. 121
[216] K.-M. Yong, G.-S. Poo, and T.-H. Cheng. Optimal placement of multicast and
wavelength converting nodes in multicast optical virtual private network. XomB
putzr Xommunixvtion“, 29(12):2169–2180, 2006. 137
[217] S. Yu, S.-C. Lee, O. Ansell, and R. Varrazza. Lossless optical packet multicast
using active vertical coupler based optical crosspoint switch matrix. Journvl of
aightfivvz izxhnology, 23(10):2984–2992, 2005. 136
[218] G. S. Zervas, M. D. Leenheer, L. Sadeghioon, D. Klonidis, Y. Qin, R. Nejabati,
D. Simeonidou, C. Develder, B. Dhoedt, and P. Demeester. Multi-granular op-
tical cross-connect: Design, analysis, and demonstration. IEEEDdhV Journvl of
dptixvl Xommunixvtion“ vny cztfiorking, 1(1):69–84, 2009. 46, 47, 123
[219] B. Zhai, D. Blaauw, D. Sylvester, and K. Flautner. Theoretical and practical
limits of dynamic voltage scaling. In DVX, San Diego, CA, June 2004. 36
[220] Y. Zhang, P. Chowdhury, M. Tornatore, and B. Mukherjee. Energy efficiency
in telecom optical networks. IEEE Xommunixvtion“ hurvzy“ vny iutorivl“,
12(4):441–458, 2010. 44, 107, 110, 121
[221] Y. Zhou and G.-S. Poo. Optical multicast over wavelength-routed WDM net-
works: A survey. dptixvl hfiitxhing vny cztfiorking, 2(3):176–197, 2005. 138
[222] Y. Zhou and G.-S. Poo. Multicast wavelength assignment for sparse wavelength
conversion in WDM networks. In IcFdXdb, Barcelona, Spain, Apr. 2006. 137
[223] Y. Zhu, Y. Jin, W. Son, W. Guo, W. Hu, W.-D. Zhong, and M.-Y. Wu. Multicast
flow aggregation in IP over optical networks. IEEE Journvl on hzlzxtzy Vrzv“ in
Xommunixvtion“, 25(5):1011–1021, 2007. 138, 139
[224] Y. Zhu and J. Jue. Multi-class flow aggregation for IPTV content delivery in IP
over optical core networks. Journvl of aightfivvz izxhnology, 27(12):1891–1903,
2009. 138, 139