Performance Analysis of Adaptive Algorithms for Gaussian and Non-Gaussian Noise Cancellation

Adaptive filtering is an extremely active area and practical field of research in signal processing. Noise cancellation using adaptive filtering is the effective technique of estimating original signal from signals corrupted by noise. Almost all adaptive algorithms which belong to different classes of adaptive filters are compared. Four criteria including GSNR, misadjustment, rate of convergence, robustness are used for performance comparison. For noise environment, both Gaussian noise and non-Gaussian noise are considered. Through comparison, it can be concluded that RLS are the best choice for every situation in terms of four criteria and NLMS, RLS are only applicable to both noise environment and other algorithms such as LMS, ENSSLMS, RVSSLMS can find their efficiency in Gaussian noise situations by adjusting their parameter appropriately.