A design technique for nonuniform quantizer in PCM generation

This paper is devoted to the use of nonuniform quantizer in order to reduce quantization noise in Communication Systems. The quantizer is designed to minimize the Mean Square quantizing Error (MSE) for a given probability density of the signal distribution. The addition of a suitable nonlinear reference waveform to the constant slope staircase function introduces the desired nonlinearity in quantizer characteristic. In the receiver also, similar process is used to decode the signal. The parameters which characterize the reference waveform are optimized to achieve the minimum MSE. Such a quantizer can profitably be used for Laplacian, Gamma, and Gaussian signals. The improvement of the signal-to-noise ratio (SNR) is shown to be considerably better than that of is law quantizer. The performance for a Laplacian distribution is investigated in detail to highlight the utility of the nonuniform quantizer.