Performance Improvements in Sub-Band Coding Using the Proposed ADM

Abstract

This paper presents a method for improving the signal-to-noise ratio (SNR) of adaptive delta modulators (ADM) with one-bit quantization. An ADM with pre-processed signals is proposed here. For example large variations in the input signal require larger step sizes. If step size is not increased, it leads to slope overload error. The large variations in the input signal are removed /smoothened by a suitable pre-processing method, one of which is using an integrator which can smoothen the rapid changes. The SONG and `improved¿ modified ABATE algorithm is used for the adaptation of step size. A speech coding system based on sub-band coding [1] with the pre-processed signals is also described here (two-channel and four-channel). Simulations have confirmed that the ADM with pre-processing leads to a high dynamic range performance and signal-to-noise ratio (SNR) compared to the conventional ADM.