Computation with phase oscillators: An oscillatory perceptron model

Abstract

We present a perceptron model with processing units consisting of coupled phase oscillators. The processing units are able to compute the input signals through a high order synapse mechanism. We show how a network of these elements can be used in analogy to the classical multilayer feedforward neural network. The main characteristics of the classical multilayer perceptron model are conserved, as for example, the backpropagation algorithm for learning. This model of coupled phase oscillators can be seen as a generic study in order to use different kind of oscillators for computational tasks.