Multiple-Access Scheme Optimisation for Artificial Neuronal Networks

Nanoscale communication will expand the scope of nanotechnology and bring new applications to the future world. Among the different means for nanoscale communication, artificial neuronal networks are a novel paradigm. The aim of this paper is to find the optimal multiple-access scheme with the objectives of maximising the number of parallel packages and minimising the firing time difference of sensors in different network sizes. This is achieved via the use of genetic algorithm. Furthermore, the performance of the multiple-access scheme in different neuron densities and collision rates is compared in terms of average packet delay, overall throughput, and energy consumption.