Efficient Programmable Random Variate Generation Accelerator from Sensor Noise
Accepted version
Peer-reviewed
Repository URI
Repository DOI
Change log
Authors
Meech, JT
Stanley-Marbell, Phillip https://orcid.org/0000-0001-7752-2083
Abstract
We introduce a method for non-uniform random number generation based on sampling a physical process in a controlled environment. We demonstrate one proof-of-concept implementation of the method that reduces the error of Monte Carlo integration of a univariate Gaussian by 1068 times while doubling the speed of the Monte Carlo simulation. We show that the supply voltage and temperature of the physical process must be controlled to prevent the mean and standard deviation of the random number generator from drifting.
Description
Keywords
C++ languages, Generators, Gaussian distribution, Temperature measurement, Monte Carlo methods, Field programmable gate arrays, Accelerometers, Bayesian, inference, noise, nonuniform, random, sensor
Journal Title
IEEE Embedded Systems Letters
Conference Name
Journal ISSN
1943-0663
1943-0671
1943-0671
Volume Title
Publisher
Institute of Electrical and Electronics Engineers (IEEE)
Publisher DOI
Rights
All rights reserved
Sponsorship
Alan Turing Institute (EP/N510129/1)
Engineering and Physical Sciences Research Council (EP/R022534/1)
Royal Society (RG170136)
Engineering and Physical Sciences Research Council (EP/L015889/1)
EPSRC (EP/V004654/1)
Engineering and Physical Sciences Research Council (EP/R022534/1)
Royal Society (RG170136)
Engineering and Physical Sciences Research Council (EP/L015889/1)
EPSRC (EP/V004654/1)
Alan Turing Institute award: TU/B/000096
EPSRC grants: EP/N510129/1, EP/R022534/1, EP/V004654/1 and EP/L015889/1