GSC Adaptive Beamforming Using Fast NLMS Algorithm for Speech Enhancement

This paper presents speech enhancement using Generalized Side lobe Canceller (GSC) beamforming with Fast Normalized Least Mean Square (FNLMS) adaptive algorithm under various real-time noisy conditions. GSC beamforming provides spatial filtering to reduce the noise in the desired signal. In this paper, fast convergence and low complexity, FNLMS adaptive algorithm is implemented and is applied to GSC structure to have enhanced speech at the output. The performance of the proposed GSC-FNLMS is compared with existing GSC-Least Mean Square (LMS) and Normalized LMS (NLMS) algorithms under various noisy conditions. GSC-FNLMS is evaluated using the performance metrics like Signal to Noise Ratio (SNR), Perceptual Evaluation of Speech Quality (PESQ) and Log Likelihood Ratio (LLR). An enhanced speech with improved performance is achieved by GSC-FNLMS compared to GSC-LMS, and GSC-NLMS algorithms.