Angle of Arrival Estimation for Saturated Acoustic Signals

Abstract

Abstract : The U.S. Army Research Laboratory (ARL) has developed a tetrahedral microphone array system for acoustic-based target detection and localization. Time-domain data can be used to obtain differential time of arrival (DTOA) and angle of arrival (AOA) estimates. Use of multiple arrays allows for triangulation. However, limitations in acoustic detection hardware, such as microphone and analog-to-digital converter (ADC) dynamic range can produce inaccurate results for large amplitude signals. Saturation can occur due to close proximity to a large transient event, which can render target localization difficult with many standard algorithms. Our goal is to develop an algorithm to detect transient events that saturate an acoustic system and estimate the AOA. This is done in three parts detection, window selection, and AOA estimation. We have achieved a 100% detection rates with no false alarms for an ARL tetrahedral microphone array located approximately 60 m from the launch position of a large-caliber, indirect fire weapons system. After outlier removal, the average absolute deviations of the AOA errors were within 5 of the true angle.