Study Of Anti-interference Ability Of Scale-free Spiking Neural Network Based On Synaptic Plasticity

With the progress and development of science and technology,the degree of automation and information in human society continues to improve,resulting in the complex electromagnetic environment become worse.However,traditional protection methods are difficult to protect electronic systems under complex electromagnetic environment.The organism under the regulation of nervous system has advantages of self-organization,self-adaption and disturbance rejection.Under the background of electromagnetic bionic protection,a scale-free spiking neural network with small-world properties is constructed.The anti-interference ability of the neural network under Gauss,impulse,AC electric field and neuron damage is studied,which lays a theoretical foundation for improving the protection ability of electronic systems in complex electromagnetic environments.The main work of this paper is as follows:(1)Based on Izhikevich neuron model and synaptic plasticity of excitatory and inhibitory synapses,a scale-free spiking neural network with small-world properties is constructed and the dynamic evolution process is studied.Synaptic regulation process,firing state and complex network characteristics of the neural network are analyzed.The experimental results show that under the regulation of synaptic plasticity,the firing rate of neurons gradually decreases and tends to be stable,and the initial firing synchronization is better,but the synchronization becomes worse as the simulation proceeds.The local information transmission efficiency is stable,the global information transmission efficiency is reduced and tends to be stable,and the small world attributes are relatively stable.The regulation of synapses is based on the firing time of neurons,and the regulation of synapses will affect the firing of neurons and complex characteristics of networks.The three influence each other to realize the dynamic evolution process of the network together.(2)The anti-interference ability of scale-free spiking neural network under Gauss,impulse and AC electric field is studied.The influence of interference on firing rate and the correlation between membrane potential of neurons before and after interference are analyzed.The experimental results show that a certain intensity of Gauss,impulse,AC electric field has little effect on firing rate and the correlation between membrane potential of neurons is high.The dynamic regulation process of synaptic plasticity weights is significantly related to the anti-interference performance of the network.The scale-free spiking neural network in this paper has a certain ability to resist Gauss,impulse and AC electric field interference under the regulation of synaptic plasticity.The range of anti-interference is discussed.(3)The anti-interference ability of scale-free spiking neural network under neuron damage is studied and is compared with that of small-world spiking neural network.The firing rate and the correlation between membrane potential of neurons are analyzed.The results are verified from the perspective of synaptic regulation and topological changes.The experimental results show that the scale-free spiking neural network has better anti-interference ability under random attack.The range of anti-interference is discussed.Compared to attacking intermediate and low degree nodes and random attack,the anti-interference of the scale-free spiking neural network is poor when the high degree nodes are attacked.The network still has certain anti-disturbance ability and the range of anti-interference is discussed.The dynamic regulation process of synaptic plasticity weights is significantly related to the anti-interference performance of the network.When the nodes are random attacked,the anti-interference of the scale-free spiking neural network is better than the small-world spiking neural network;when the high degree nodes are attacked,the anti-interference of the small-world spiking neural network is better than the scale-free spiking neural network.