Maximum a Posteriori (map) Multitarget Tracking for Broadband Aeroacoustic Sources

Acoustic sensors offer the potential to detect, track, and classify enemy ground targets such as tanks, trucks, and other military vehicles. Traditional multitarget tracking techniques partition the track estimation operation into two isolated processes: directionof-arrival (DOA) estimation from array snapshot data, followed by track estimation from the DOA estimates. In this paper, a multiple target track estimation method that operates directly from broadband array data is presented. The maximum a-posteriori (MAP) estimator for contact states is derived for temporally uncorrelated broadband target signals and uncorrelated target tracks. This technique extends the MAP multitarget tracking approach developed for narrowband signals in [1]. The track estimator is an iterative algorithm employing a nonlinear programming penalty method in conjunction with an alternating maximization algorithm for obtaining penalized maximum likelihood (PML) DOA estimates. The penalty function couples the DOA estimates from the PML algorithm to the tracker as synthetic measurements, eliminating the data association step of traditional multitarget tracking approaches. It also creates a feedback mechanism to enhance the DOA estimation process. The algorithm is derived as a batch method. A sequential implementation obtained by stepping through the data in short batches is applied to acoustic array time series data from field tests conducted by the Army Research Laboratory.