PROPAGATING COALITIONS IN NETWORKS OF NONLINEAR OSCILLATORS

抄録

A brain can perform pattern recognition tasks that are not yet possible for an electronic computer. We continue here our investigation of electronic circuits that are inspired by knowledge of structures in a brain. These circuits are oscillators that are motivated by principles of neuroscience, but yet are constructible as micro circuits, and possibly as nano-circuits. Populations of such oscillators can exhibit patterns in their output frequencies. This may be in response to an external signal being applied across the population or in response to internal waves that propagate through the population. We have investigated the former phenomenon where coalitions of oscillators, classified by their output frequency, form in response to a common driving signal. The resulting patterns have been used to characterize the input signal and they serve as a basis for comparison of signals and other pattern recognition tasks. In this paper, we investigate pattern formation that results from the propagation of synchronized activity waves within the population of oscillators. The novelty here is in the model: We derive a nonlinear wave equation to describe networks of oscillators, and simulate solutions to demonstrate some basic results.