Self-Learning Blind Wireless Channel Identifier Using Adaptive Algorithms

Abstract

The adaptive filter adapts its coefficients according to the error between its output and the received signal samples . The adaptive filters have various applications such as channel identification , equalization , noise cancelation and prediction. Blind wireless channel identification is the scope of this thesis. In this thesis we propose blind adaptive algorithms (least mean squares (LMS) , normalized least mean squares (NLMS) , recursive least squares (RLS) , constant modulus algorithm (CMA) and Sato) to address the problem of blind adaptive wireless channel identification and provide good adaptive filtering performance. Expressions to update the adaptive filter coefficients are derived. We develop these algorithms to provide comparisons in the convergence speed , mean squared errors (MSE) and complexity. Different filter tap-weights and different signal to noise ratio (SNR) values are used to provide comparable results .The performance of the proposed algorithms are analyzed and simulation results showed that the RLS algorithm has fastest convergence and less MSE value as compared with other proposed adaptive algorithms . Finally , the results showed that as the adaptive filter length increases , the convergence of its coefficients to the optimum values becomes slower.