The research on the stochastic resonance based of feedback FitzHugh-Nagumo neural network

The research on stochastic resonance phenomenon of neuron had shown the important theoretical significance and application value of the weak signal detection. The robustness performed not very well during the process of the weak signal detection, which based on the stochastic resonance of the traditional FitzHugh-Nagumo (FHN) neuron model. The addition of feedback loop which achieved the reaction formation from the model response to the input layer, could improve the performance of the weak signal detection. Comparative analyses of the traditional and improved FHN neural network were taken by combining with the spike frequency and amplitude. The results show that the responses of stochastic resonance based on feedback FHN neural network possess better performance and stability during a certain range of noise intensity. Thus, the stochastic resonance of this improved feedback FHN network can be more perfectly applied to the weak signal detection and transmission.