On the Performance Limits of Array-Based Acoustic Source Localization

Accurate localization of acoustic sources is crucial for detecting and identifying sound events. However, in practical scenarios, the performance of acoustic source localization can be affected by various factors such as noise, sensor placement deviations, number of sensors, and sensor array layout. This article aims to investigate the achievable localization performance limits of array-based acoustic source localization schemes, identify the critical factors that influence the actual localization performance, and provide recommendations and optimizations for array layouts in different scenarios. To achieve this, we present a Cramér–Rao lower bound (CRLB) analysis to reveal the theoretically achievable localization performance limits with noise levels, sensor placement deviations, number of sensors, and array layout. Our analysis highlights the significant impact of sensor placement deviation on the array-based acoustic localization performance. Considering various acoustic source localization scenarios, we validate the effectiveness of the proposed method in both 2-D and 3-D spaces. Based on this, we further recommend several array layouts for acoustic source localization. Furthermore, we uncover the viable region with guaranteed localization performance, directionality phenomena, and perception capability of arrays, which aid in optimizing array layouts in practical applications. In summary, this study provides valuable insights into the critical factors affecting acoustic source localization performance and suggests optimal array layouts for different scenarios, which can significantly improve the accuracy of acoustic event detection in practical applications.