Optimum Adaptive Filtering Algorithms To Improve The Performance Of Communication System

Abstract

quotIn communication system, signals are distorted by additive back ground noise while propagating through the media. At the receiving end, listener will not able to receive intended information due to the presence of this additive noise. It is important to estimate newlinethe intended information from the corrupted one for reliable communication system. Filters are systems in signal processing which are used to estimate intended information from the corrupted signal. In an ideal environment classical filters may be used to recover newlinethe signal and these are rarely optimum to produce the best estimate of signals. Optimum filters like Wiener filter is used to produce the best estimate of the signal, but this filter is optimum only when the statistical characteristics of the input signal matches with the prior information of the signal on which the filter has been designed. If the prior information of the signal is unknown then, adaptive filters are used to estimate the signal. An adaptive filter is one whose transfer function is adjusted according to an optimized algorithm driven by an error signal. The Least mean square (LMS) algorithm is the most popular algorithm in adaptive signal processing. Its performance depends primarily on the value of constant feedback parameter and#956;. A large value of feedback parameter and#956; leads to fast rate of convergence with increased misadjustment and smaller the value of and#956; leads to newlinea slow rate of convergence with reduced misadjustment. For further improvement, many LMS based Variable Step Size (VSS) algorithms have been proposed. In these algorithms the value of convergence parameter varies with respect to time to achieve faster rate of convergence and small misadjustment with increased computational complexity. newlineHowever, the performance of these algorithms is sub-optimal and they have their own drawbacks. This provides further avenues for research in the development of optimum LMS based VSS algorithms to improve the performance of communication system. newline