Research On Stochastic Resonance Of Complex Neural Networks Stimulated By Pulse Signals

With the rapid development of society,ageing population increases day by day.Environmental pollution caused by industrial production directly leads to more and more diseases,and to adapt to the development of times,people are forced to work overtime for a long time,causing a series of health problems.Most of those problems are in the nervous system such as the sense of touch,hearing and visual perception.A large number of studies have shown that noise widely exists in the nervous system and can enhance the information processing ability of the neurons,which is stochastic resonance.This phenomenon reflects that noise can enhance the awareness of neural networks to external stimuli,and to some extent,it also indicates a kind of information processing technology in improving people's quality of life.In this paper,we select cross-correlation coefficient as an evaluation index and explore stochastic resonance in random uniform sparse networks and exponential networks composed of integrate-and-fire neuron models.It is noted that the coupling networks are driven together by input discrete pulse signal and the internal noise.First,we choose Gaussian white noise and uniform distribution noise to simulate internal noises in neurons,and discrete pulse signal whose time interval obeys Poisson distribution as input signal of two kinds of coupling networks.Second,we analyze effects of the threshold voltage,the noise type and the structure of network on the cross-correlation coefficient of the output-input firing rates in different amounts of sensory neurons.It is proved that,as the internal noise intensity increases,the neurons have a better response to the input signal,wherein the cross-correlation coefficient can reach a maximum value in a certain range of noise intensity.These results indicate that stochastic resonance plays a critical role in the field of the neural information processing.Finally,we present a set of algorithms for constructing simulation models of the complex network automatically.By using the Simulink platform we can connect nodes in the complex network with directed links.And the coupling coefficients of subsystems form the weighed matrix,and different network topologies can be established by assigning the corresponding matrix element values.The significant improvement of efficiency provides the basis for further research on the evolution nature of complex network dynamics.