Design and Analysis of Cascaded LMS Adaptive Filters for Noise Cancellation

Adaptive filters have become active research area in the field of communication system. This paper investigates the innovative concept of adaptive noise cancellation (ANC) using cascaded form of least-mean-square (LMS) adaptive filters. The concept of cascading and its algorithm for real-time LMS-ANC are also described in detail. The model of the cascaded LMS-ANC is designed and simulated on MATLAB Simulink. The simulation model gives variation in the distinct signals of LMS-ANC like error, output and weights at various LMS filter parameters. The proposed algorithm utilizes two adaptive filters to estimate gradients accurately and results in good adaptation and performance. The objective of the present investigation is to provide solution in order to improve the performance of noise canceller in terms of filter parameters. The results are obtained with the help of adaptive algorithm with variable step size and different initial weight of filters which provides high convergence speed of error signal. This paper also includes the derivation for the convergence rate at different conditions and concludes that cascaded LMS-ANC results in higher convergence rate and better output signal as compared to single LMS-ANC. Higher signal-to-noise ratio for cascaded system is obtained for cascaded LMS-ANC as compared to that of single LMS-ANC system.