Design of Nonuniform Dead-zone Quantizer with Low Number of Quantization Levels for the Laplacian Source

This paper proposes a nonuniform dead-zone quantizer for quantization of the speech signal, modeled with the Laplacian probability density function. It is consisted of the inner dead-zone part and the outer nonuniform Lloyd-Max’s part. The numerical results provided in the paper show that, for the fixed value of source entropy, the proposed quantizer provides the gain in the quality of the quantized signal in comparison to other quantizer models with low number of quantization levels. Accordingly, one can believe that the proposed quantizer will find its way toward the practical implementation, especially in speech quantization.