A novel time-event-driven algorithm for simulating spiking neural networks based on circular array

Abstract

Abstract The computing of synaptic currents occupies a major part of computational cost when simulating a large scale spiking neural network. Based on the observation that the probability of a neuron receiving at least one spike from any synapses during a very tiny simulation time step is very small, we propose a time-driven algorithm corrected by an event-driven process (a hybrid time-event-driven algorithm) which consists of two procedures of computation. In the first procedure of the synaptic current computation, we suppose that the neuron in question receives no spike during the simulation time step, and thereby propose a time-driven method of joint decay process to reduce the computational complexity of the synaptic current. In the second procedure of the computation, we suppose that the neuron in question receives spikes during the simulation time step, and propose an event-driven local correction process to correct the total synaptic current that is calculated in the first procedure of the computation. We design a data structure of circular two-dimensional array for storing both conductance coefficients related with presynaptic neurons and correcting conductance related with postsynaptic neurons. Furthermore, in order to realize the local correction process quickly and effectively, we propose a new event-processing method to realize the local correction process based on the data structure of circular two-dimensional array. By theoretically comparing with that of traditional time-driven algorithm, it is found that the proposed time-event-driven algorithm reduces computational cost of synaptic current substantially. The simulation results further show the efficiency of the proposed algorithm.