Abstract
Molecular communication is an emerging communication technology for applications requiring nanoscale networks. Transferring vital information about external and internal conditions of the body through the nervous system is an important type of intra-body molecular nanonetworks. Thus, investigating the performance of such systems from the communication theoretic perspective gives us insight on the limitation of neuro-spike communication and ways to design artificial neural systems. In this letter, we study the performance of the neuro-spike communication under different stochastic impairments such as axonal shot noise, synaptic noise, and random vesicle release. The objective is to optimally detect the spikes at the receiving neuron. Since several uncertainties occur under each hypothesis, composite hypothesis is employed to find the optimum detection policy. Furthermore, we obtain closed-form solutions for the optimal detector and derive the binary decision error at the postsynaptic terminal.