Intrinsic Oscillations of Neural Networks
Howard Hua Yang · 2005
The asymmetric network is a more plausible model for biological networks. Its dynamics is affluent but its theory is poor. Oscillation theory including the limit cycle theory for high dimensional nonlinear systems is usually un-feasible in application. An oscillation result is presented in this paper for networks with ring structure. Without self-connections, a neural oscillator can be built with at least three neurons. Inhibitory neurons are important in generating intrinsic oscillations when the external input is absent. More complicated oscillator can be formed by coupling the neural rings. 1 Introduction The papers to be cited: [1, 2, 6, 7, 9] The coherence or synchrony of the oscillations between different parts of the cortex plays an important role in binding features detected at separated cortex areas to perceive objects[4, 5, 8]. An oscillatory network provides a potential solution for pattern segmentation[5, 7]. The oscillations in neural responses have been observed...