Abstract—In this paper the eminence of the Fast Transversal Least Mean Squares (FT-LMS) algorithm over LMS and RLS algorithms is provided. This algorithm is designed to provide similar performance to the standard LMS algorithm while reducing the computation order. Finite precision effects are also briefly discussed. Simulations are performed with these algorithms to compare both the computational burden as well as the performance for adaptive noise cancellation. Simulations show that FT-LMS offers comparable performance with respect to the standard RLS and LMS in addition to a large reduction in computation time for higher order filters.
Index Terms—Adaptive filters, LMS, RLS, FT-LMS.
D. Hari Hara Santosh is with MVGR College of Engineering, India (e-mail: santosh_ece2005@yahoo.com).
Cite: D. Hari Hara Santosh, Samalla Aditya, K. Sarat Chandra, and P. Surya Prasad, "Performance Analysis of Noise Cancellation in Speech Signals Using LMS, FT-LMS and RLS Algorithms," International Journal of Modeling and Optimization vol. 2, no. 6, pp. 667-671, 2012.
Copyright © 2008-2024. International Journal of Modeling and Optimization. All rights reserved.
E-mail: ijmo@iacsitp.com