Implementation of Spiking Neural Network with Wireless Communications

Abstract

This paper proposes and implements the Spiking Neural Network (SNN) with radio-frequency wireless communications. The implemented network could obtain the XOR function through reinforcement learning. By applying the wireless communication for Internet of Things to the SNN, the SNN works with sufficient communication distance and low power consumptions for not only the line of sight environment but also the non-line of sight one. Additionally, it is unnecessary to consider communication directivity and obstacles for constructing the networks. The experimental results showed the extensibility and the scalability of the implemented system in this paper.