Improved active noise control algorithm based on the convex combination method

Noise can be harmful in certain situations, which leads to the need of controlling it. In general, a particular method of active noise control basically consists of creating a destructive overlap between noise and a reference signal (both with the same frequency and amplitude but with an opposite phase). Different algorithms can be used for this purpose, each of them having distinct features, like convergence, velocity, and steady state attenuation level. The convex combination method for active noise control (C-ANC) has been used successfully as it combines different characteristics of most algorithms found in literature. This method combines two distinct active noise control algorithms into a single one. For instance, it can use algorithms such as FXLMF, which has good attenuation velocity, and the FXLMS algorithm, which has a significant attenuation level. This work proposes a modified convex combination algorithm, named MC-FXLMS-F, in order to maximize the overall attenuation level. Modification of the convex combination method was implemented using the FXLMS-F algorithm. The new proposed algorithm was compared to others found in literature, such as the FXLMS, FXLMS-F, and C-FXLMS-F algorithms. The modified convex combination algorithm (MC-FXLMS-F) showed improvement in both convergence speed and attenuation level. It achieved the best performance among the most popular existing algorithms.