Seismo-acoustic Vector Detection and Tracking of Ground Vehicles

Abstract:

The Army Research Laboratory is undertaking an investigation to combine acoustic and seismic vector sensors, in conjunction with traditional infrasound arrays that use omnidirectional microbarometers, to detect and track ground convoys at standoff. The focus of this study is to improve the fidelity of the track with a sparse network, and explore its applicability against variations in topography, geology, speed and standoff. Combining low-frequency seismic and acoustic sensing, we demonstrate a novel capability to extend the standoff range while maintaining some of the characteristics of the target signature. Towards this end, we carry out a number of field tests where multiple vehicles and convoys that include multiple types of vehicles, and with known traverses. Using a detection algorithm that leverages both the vehicle dynamics and the signal propagation with distance, we explore this new capability to detect and track the vehicles as a function of local geology and propagation conditions. Secondly, both acoustic and seismic vector sensors have the ability to estimate both the arrival time and the azimuth to the source that lead to smaller sensor footprint, as compared to an array. By combining multimodal sensing with novel target tracking techniques, we can reduce the impact of clutter while maintaining a robustness of our detection and tracking algorithms. As expected, interference from manmade clutter signals, e.g., unknown vehicles, aircraft, etc., has a significant impact on the classification of the convoys. We will present a comparison of the acoustic and seismic vector sensing in tracking ground vehicles.

Recently, ARL has developed a novel deployable acoustic array that is suitable for detecting and tracking narrowband tones at standoff. This broadband spiral array is applicable over a range of frequencies and azimuths, and is ideal for wide area monitoring with unknown background clutter. Though the time varying tonal signals from vehicles presents a challenge for frequency-domain detection, we show that a combination of broadband and narrowband signals can improve its robustness. We will discuss the applicability of spiral arrays for vehicle detection and tracking.