An Improved Time Delay Estimation Algorithm in Microphone Array Sound Source Localization

In order to enhance the precision of sound source localization in microphone arrays operating within low signal-to-noise ratio (SNR) environments, an improved time delay estimation algorithm is proposed based on the generalized cross-correlation (GCC) algorithm. The algorithm suppresses noise interference on the signal, and increases the relative values of the main and side peaks of the cross-correlation function, thereby modifying the accuracy of time delay estimation. A simulation experiment involving time delay estimation and sound source localization based on a six-element cross-microphone array is conducted in MATLAB. A comparative analysis of the time delay estimation performance between the proposed algorithm, the GCC algorithm, and the quadratic correlation algorithm is conducted. Furthermore, the sound source localization effect of the improved algorithm is simulated under low SNR conditions. The simulation results reveal that the proposed modified algorithm, compared to the GCC and quadratic correlation algorithms, sharpens the peak of the cross-correlation function, significantly enhances the accuracy of time delay estimation, and exhibits robust noise resistance and smaller sound source localization errors in low SNR environments.