Internet of Nano Things (IoNT) is an emerging technology, which aims at extending the connectivity into nanoscale and biological environments with collaborative networks of artificial nanomachines and biological entities integrated into the Internet. To enable the IoNT and its groundbreaking applications, such as real-time intrabody health monitoring, it is imperative to devise nanoscale communication techniques with low-complexity transceiver architectures. Bio-inspired molecular communications (MC), which uses molecules to transfer information, is the most promising technique to realise IoNT due to its inherent biocompatibility and reliability in physiologically-relevant environments.

Despite the substantial body of work concerning MC, the implications of an interface between MC channel and practical MC transceiver architectures are largely neglected, leading to a major gap between theory and practice. As the first step to remove this discrepancy, in this thesis, I develop a realistic analytical ICT model for microfluidic MC with surface-based receivers as a convection-diffusion-reaction system.

In the second part, I focus on biological MC receivers, which can be implemented in living cells using synthetic biology tools. In this direction, I theoretically develop low-complexity and reliable MC detection methods exploiting the various statistics of the stochastic ligand-receptor interactions at the membrane of biological MC receivers. The estimation and detection theoretical analysis of these detection methods demonstrate that even single type of receptors can provide sufficient statistics to overcome the receptor saturation problem, cope with the interference of non-cognate molecules, and simultaneously sense the concentration of multiple types of ligands. I also propose synthetic receptor designs for the transduction of decision statistics into a representation by concentration of intracellular molecules, and design chemical reaction networks performing decoding with intracellular reactions.

Finally, I fabricate a micro/nanoscale MC receiver based on graphene field-effect transistor biosensors and perform its ICT characterisation in a custom-designed microfluidic MC system with the information encoded into the concentration of DNAs. This experimental platform is the first practical demonstration of micro/nanoscale MC, and can serve as a testbed for developing realistic MC methods.