Studying the dynamics of memristive synapses in spiking neuromorphic systems

The paper describes the numerical study of a simple spiking neuromorphic system comprising three neurons. We consider the Hodgkin-Huxley neuron model represented by equivalent electrical circuit to simulate the dynamical behavior of ionic channels in a cell membrane. The neurons in the system are connected to each other via synapses represented by memristive elements. The applied model of memristive element with two adaptive thresholds allows us to achieve several basic properties of the spike-timing-dependent plasticity (STDP) of biological synapses. The corresponding mathematical model is given as a system of ordinary differential equations. As the experiment findings, we present the results of dynamical analysis of the three-neuron system carried out with extrapolation methods of numerical integration.