THE APPLICATION OF NONLINEAR STRUCTURES TO THE RECONSTRUCTION OF BINARY SIGNALS

This paper considers the problem of reconstructing digital signals which have been passed through a dispersive channel and corrupted with additive noise, a problem which is of considerable importance in communications. The problems encountered by linear equalizers under adverse conditions on the signal-to-noise ratio and channel phase are described, and by considering the equalization problem as a geometric classification problem we demonstrate how these difficulties can be overcome by utilizing nonlinear classifiers as channel equalizers. The manner in which neural networks can be utilized as adaptive channel equalizers is described and simulation results are presented which suggest that the neural network equalizers offer a performance which exceeds that of the linear structures, particularly in the high-noise environment.