Reference Weighted Filtered-x LMS Algorithm for Active Control of Impulsive Noise

This paper proposes a reference-weighted filtered-x least mean squares (RWFxLMS) algorithm for active control of impulsive noise. The paper briefly discusses the existing algorithms that inspired the development of the proposed algorithm. Thereafter, the concept of the proposed algorithm is described. The RWFxLMS algorithm uses a quadratic function of the reference signal obtained by curve-fitting the probability distribution of an impulsive signal to weight the step-size at that instant. Analysis related to the convergence property of the proposed algorithm is made, and its computational complexity is calculated. Its performance against a variety of impulsive noise signals - random impulsive with varying degree of impulsiveness and repetitive impulses - is studied. Lastly, the algorithm is compared with Y. Zhou's FxGSNLMS algorithm in terms of computational complexity and convergence properties.