Research On Sound Orientation Spike Coding Mechanism

Recently, Spiking Neural Networks(SNNs) as the next generation of Artificial Neural Networks is researched by more and more people.The main difference between SNNs and traditional Neural Networks is the information encoding type . The encoding type of SNNs is temporal, based on the timing of single spikes, instead of average firing frequency. So SNNs model fits to analyze the time-structured problems with its powful computation capability and simple structure.Therefore, we simulate the information process of the biosonar neural system to study the biosonar mechanism.The computational neuroscience is discussed in the order space which is only to compare the relative values instead of computing the absolute values. During this paper, we establish the H-H spike neural network model to process the input signals.The output strings of spikes which contain tempo-space information transform into strings of symbols according to circle mapping and symbol-dynamics. Applying this model, we discuss and explain the mechanism of biosonar system how to estimate the distance of object, coming direction, velocity and the tendency of the body's movement.