Study Of Cortex Networks Synchronized Activity Based On Spiking Neurons Synaptic Plasticity

With the development of information era,electromagnetic environment becomes more and more complex.Finding new methods to improve the robustness of electronic systems in complex electromagnetic environment has been very important.The brain is capable of exhibiting good adaptive robustness in an unstable external environment,which the traditional electronic systems cannot match.Self-organization immunity mechanism and the information processing mechanism of the brain are consistent.Which are closely related to the synchronization of neurons spiking activities.Based on the concept of electromagnetic bionics protection,in this thesis,we focus on the relationship between synchronization of neuron firing activity and characteristics of homogeneous and heterogeneous feedforward cortex networks,such as synaptic connection and noise excitation.Which provide certain theoretical support for the research of bionics.Main work is as follows:(1)By studying the spiking characteristics of cortical networks under different connection probabilities.The inference impact of network connection on neuron spiking synchronous activity is analyzed.Established the cortex network model of different sizes.Effects of neurons number and network topology on synchronizing characteristics and frequency characteristics are studied.Analyzed the relationship between synchronous degree of neuron spiking and connection probability as well as network size.The results show that Fully connection mode of the neural network has better synchronization spiking characteristics.The synchronous spiking degree under the rest network connection related to the network size.(2)Through analyzing the spiking characteristics under noise interference and node,the robustness of adaptive neural network is illustrated.By studying the effect of noise stimulation and stimulation intensity on neural network spiking characteristics.We found that noise can induce synchronous spiking activities among uncorrelated neurons.And noise also restrain synchronous spiking activities among correlated neurons.Finally nonlinear synchronous spiking activities of the cortex network tend to be more consistent.Conduct simulation experiment by damaging nodes with different connection probability.The results show that the established feedforward cortex network model has good self-adaptive robust properties.(3)Through the establishment of heterogeneous network model feed-forward cortex,the effects of inhibitory neurons and synaptic plasticity on spiking characteristics of network is analyzed.Based on the excitatory and inhibitory synaptic plasticity,build a heterogeneous feed-forward cortex network is built.Comparing the effects of different synaptic plasticity mechanisms on spiking characteristics of the cortical network.The results show that inhibitory neurons and synaptic plasticity can reduce the spiking correlation of cortex network.