Synthesis of model of hardware realization of izhikevich model of biological neuron on the basis of FPGA

Abstract

For years of researches and developments in the field of neurotechnologies a big variety of the artificial neural networks somewhat reminding work of biological neural systems was created. In article the neuron of spiking neural network is considered. Priority fields of use of neural networks of this type: spatial navigation and analysis of the environment; analysis of smells and dynamic graphic and sound information; neuroprosthetics and neurointerfaces. Having analysed and having estimated functionality of various models of neurons for hardware realization of neural network, Izhikevich's model was chosen. This model is biologically realistic and has the lowered requirements to computing resources. The mathematical model of neuron and model of its hardware realization on the basis of FPGA is considered.