Adaptive filtering is at the core of many signal processing applications as phenomenal advances both in research and application have been made during the past three decades. The main objective of an adaptive filter is to minimize error signal at the cost of high convergence rate and reduced computational complexity. This paper presents the performance of adaptive filter algorithms for acoustic echo cancellation by eliminating the echo signal from the original signal. The adaptive filtering process is carried out by using Least Mean Square (LMS), Normalized-Least Mean Square (NLMS) and Kalman filter in a real-time non- stationary environment and its performance is measured in terms of convergence rate, Mean Square Error and Echo Return Loss Enhancement (ERLE). Simulations are carried out by using MatLab and the output results show clearly that Kalman filter converges after about 7000 iterations and outperforms LMS and NLMS algorithms with 42dB ERLE against 12dB ERLE and 20dB ERLE for LMS and NLMS respectively. Therefore, Kalman filter is more suitable for echo cancellation in a non-stationary environment.

Keywords : Acoustic Echo Cancellation, Adaptive filtering, LMS, NLMS, Kalman filter.