Simple recurrent network in real time astrocyte

Abstract

In this paper we propose simple recurrent network (SRN) and mathematical paradigm to model real time interaction of astrocyte in simplified spiking neural network (SRM0). Both Elman neural network and Jordan neural network are used in the proposed network. Elman neural network is connected to the synapse and Jordan neural network is connected to the postsynaptic neuron in tripartite synapse, the simulation results showed that whenever astrocytes were activating in a time window, more spikes will be fired for excitatory neurons. This is biologically justified since astrocytes induce excitability with variations in Ca2+ concentrations.