MATLAB Simulation of Near-Field Acoustic Source Localization Using TDOA

In this article, we explore MATLAB-based simulation techniques for near-field acoustic source localization using Time Difference of Arrival (TDOA). The implementation includes two key algorithms: the traditional Cross-Correlation (CC) method and the Phase Transform variant of Generalized Cross-Correlation (GCC-PHAT). The CC algorithm calculates time delays by finding the peak position in the cross-correlation function between microphone signals, while GCC-PHAT improves robustness in reverberant environments by whitening the signals in the frequency domain using phase information. We analyze the comparative advantages and limitations of both approaches, examining their performance under different environmental conditions. The simulation demonstrates how to adjust critical parameters such as sampling frequency, window size, and threshold settings to optimize localization accuracy. Through practical MATLAB code examples, we illustrate signal preprocessing techniques, delay estimation methods, and source position calculation using geometric triangulation. The implementation includes functions for microphone array configuration, signal correlation analysis, and TDOA-based position estimation algorithms. Finally, we discuss practical applications of these algorithms in real-world acoustic source localization scenarios, providing implementation tips for handling noise reduction, multipath effects, and computational efficiency optimization in MATLAB simulations.