A simple spike neural network can generate complex firing patterns

In this paper, we develop a double feedback neural unit for generating impulse sequence that is used in nervous systems to transmit information. One of advantages of the DFNU is in its simple structure using only double feedback loops. Simulation results show that a wide range of firing patterns from cyclic and synchronized oscillation to random firing can be observed from a randomly coupled network consisting of DFNUs by changing only one parameter of the networks. Especially, choosing a critical value within a specific range for the single parameter, synchronous dynamics with various frequency range, or chaotic firing patterns that observed widely in biological nervous system, can also be generated by the network.