An Adaptive Coding Scheme For Effective Bandwidth And Power Utilization Over Noisy Communication Channels

Abstract

Codes for communication channels are in most cases chosen on the basis of the signal to noise ratio expected on a given transmission channel. The worst possible noise condition is normally assumed in the choice of appropriate codes such that a specified minimum error shall result during transmission on the channel. However, as the noise conditions on communication channels vary from time to time, the use of a particular code that meets the specified error criteria under the assumption of worst noise condition may not result in the optimum utilization of the channel. One can exploit the variations of the noise level on the channel to increase the throughput on a fixed bandwidth channel without increasing the transmitter power by the use a coding scheme that adapts itself to the noise conditions on the channel.

An adaptive coding scheme that utilizes the different merits of more than one fixed rate codes and the noise level variations on a communication channel to achieve an increased throughput has been implemented in this work. An adaptive coding scheme using three fixed-rate BCH codes was developed and implemented on a computer. The
results obtained in transmitting a given text over a simulated channel indicate an improved performance in terms of download time and coding
gain. The use of the adaptive coding scheme results in a saving of 28% in download time and a coding gain of 2.61 dB when compared to the performance of a given reference code.