A one bit neural A/D converter

A well-known class of one-bit A/D converters are delta modulators. There are two main classes of delta modulators; linear and adaptive. In this paper an adaptive delta modulator, where adaptation is achieved with a neural network, is described. The network is of the perceptron type and consists of five inputs, one hidden layer and an output. We have simulated signal-to-noise ratio for three different types of delta modulators, and results have shown the newly proposed delta modulator has better signal to noise ratio over given dynamic range