Effectiveness Evaluation Through Z-Test in Auditory Perception of an Adaptive Noise Canceling Filter Based on Least Mean Squares Algorithm

This paper presents a comprehensive evaluation using a Z-test to assess the 1 effectiveness of an adaptive Least Mean Squares (LMS) filter driven by the Steepest Descent 2 Method (SDM). The study utilizes a male voice recording, captured in a controlled studio 3 environment, to which persistent Gaussian noise was intentionally introduced, simulating 4 real-world interference. All signal processing methods were implemented accordingly 5 in MATLAB. The adaptive filter demonstrated a significant improvement of 20 dB in 6 Signal-to-Noise Ratio (SNR) following the initial optimization of the filter parameter μ. 7 To further assess the LMS filter’s performance, an empirical experiment was conducted 8 with 30 young adults, aged between 20 and 30 years, who were tasked with qualitatively 9 distinguishing between the clean and noise-corrupted signals (blind test). The quantitative 10 analysis and statistical evaluation of the participants’ responses revealed that a significant 11 majority, specifically 80%, were able to reliably identify the noise-affected and filtered 12 signals. This outcome highlights the LMS filter’s potential—despite the slow convergence of 13 the SDM—for enhancing signal clarity in noise-contaminated environments, thus validating 14 its practical application in speech processing and noise reduction