Design and Analysis of Configurable Multilayer Spiking Neural Network

Spiking neural networks have emerged as the most biologically cogent neural networks and thus paves the way for artificially solving the real time tasks of recognition. In this paper, the design of a multiple layer spiking neuron network is proposed. For the designed spiking neural network implementation, every neuron is spiking neuron and is modeled using leaky integrate and fire. The novelty of the design is configurability in terms of figuring the layers and neurons in each layer without use of any third party neuromorphic library. The synaptic plasticity in the network varies according to spike time dependent plasticity algorithm. The results presented here are meant for multilayer spiking neural network which shows the variation of membrane potential of all neurons in each layer.