A Fast Affine Projection Algorithm Based on Matching Pursuit in Adaptive Noise Cancellation for Speech Enhancement

TL;DR

This paper introduces a fast affine projection algorithm based on matching pursuit for adaptive noise cancellation in speech enhancement, offering rapid convergence and effective noise attenuation.

Contribution

It presents a novel adaptive filtering algorithm, FAPA, that improves convergence speed and reduces complexity compared to traditional methods like LMS and RLS.

Findings

FAPA effectively attenuates noise in speech signals.

Simulation results show superior convergence speed.

The algorithm balances complexity and performance well.

Abstract

In many application of noise cancellation, the changes in signal characteristics could be quite fast. This requires the utilization of adaptive algorithms, which converge rapidly. Least Mean Squares (LMS) adaptive filters have been used in a wide range of signal processing application. The Recursive Least Squares (RLS) algorithm has established itself as the "ultimate" adaptive filtering algorithm in the sense that it is the adaptive filter exhibiting the best convergence behavior. Unfortunately, practical implementations of the algorithm are often associated with high computational complexity and/or poor numerical properties. Recently adaptive filtering was presented that was based on Matching Pursuits, have a nice tradeoff between complexity and the convergence speed. This paper describes a new approach for noise cancellation in speech enhancement using the new adaptive filtering algorithm named fast affine projection algorithm (FAPA). The simulation results demonstrate the good performance of the FAPA in attenuating the noise.