Loading...
Improved Active Noise Cancellation Using Variable Step-Size Combined Fx-LMS Algorithm
Author(s)
Kar, Asutosh
Burra, Srikanth
Shoba, S.
Vasundhara
Mladenovic, Vladimir
Goel, Pankaj
Kumar, Sanjeev
Date Issued
January 1, 2025
Type
Article
Abstract
Active noise cancellation of audio signals, which has been the subject of a substantial quantity of research in the past, remains a significant obstacle in many aspects of acoustic signal processing. In a natural environment, the system’s parameters are in constant flux. Consequently, we shall investigate the filtered-x least mean square (FxLMS) algorithm. The adaptive filter is used as a controller to generate an anti-noise signal in the feedforward FxLMS algorithm, which only addressed narrowband and broadband noise. The feedback FxLMS algorithm, on the other hand, enhances the algorithm’s convergence, but cannot eliminate the effect of residual noise’s uncorrelated disturbance. However, both the correlated and uncorrelated disturbances must be under user control. Therefore, a combination of the feedback and feed-forward FxLMS algorithms was employed, which consists of two portions for estimating the primary disturbances and the uncorrelated disturbances. A variable step size algorithm has been employed to attain a balance between system stability and control efficiency. In Akhtar’s method, an external adaptive filter was applied to the feed-forward system in order to estimate a more precise error signal and enhance the performance of the systems. These improvements will improve the algorithm’s ability to control both primary and uncorrelated noise.
Citation
Circuits Systems and Signal Processing, 44(1), 447-461, 2025