Double Direction Adaptation for Noise Reduction in Pre-Whitened Lms-Type Algorithms

The LMS algorithm suffers from its slow rate of convergence, especially for high correlated input signal. The input pre-whitening based algorithms provide better convergence rate with the price of noise enhancement. To mitigate this drawback, we present in this paper a technique, which consists on exciting the adaptive filter at both the input signal direction and the pre-whitened input direction. Hence, two different step sizes are used, they permit to improve convergence rate without enhancing the noise during steady state. A theoretical analysis of the steady state performance is presented. Simulation results are also presented to support the analysis and to compare the proposed algorithm with classical ones