Tracking the Tracker from its Passive Sonar ML-PDA Estimates

Abstract

Target motion analysis with wideband passive sonar has received much attention. Maximumlikelihood probabilistic data-association (ML-PDA) represents an asymptotically efficient estimator for deterministic target motion, and is especially well-suited for low-observable targets; the results presented here apply to situations with higher signal to noise ratio as well, including of course the situation of a deterministic target observed via "clean" measurements without false alarms or missed detections. Here we study the inverse problem, namely, how to identify the observing platform (following a "two-leg" motion model) from the results of the target estimation process, i.e. the estimated target state and the Fisher information matrix, quantities we assume an eavesdropper might intercept. We tackle the problem and we present observability properties, with supporting simulation results. Index TermsEavesdropper, Fisher information matrix, ML-PDA, nonlinear identification, platform localization, stealthy platform, wideband passive sonar. P. K.

Distributed compressed sensing based joint detection and tracking for multistatic radar system

“…JOMP and OSGA algorithms, in reconstructing a sparse vector representing the state space with high resolution. Secondly, the proposed distributed compressed sensing based joint detection and tracking algorithm is compared to the TBD algorithm [34,43], the maximum-likelihood probabilistic data association (ML-PDA) algorithm [20,10], and the maximumlikelihood probabilistic multihypothesis (ML-PMHT) algorithm [35], in tracking multiple targets (including prominent and weak targets) in a 3D Cartesian coordinate system.…”

“…The performance of the proposed distributed compressed sensing based joint detection and tracking (DCS-JDT) algorithm is further compared to the TBD algorithm [34,43], the ML-PDA algorithm [20,10], and the ML-PMHT algorithm [35]. The following metrics are evaluated to verify the performance of the algorithms: the global root mean square error (RMSE) in position and the execution time (ET).…”

Passive Sonar Target Detection Using Statistical Classifier and Adaptive Threshold

“…Sen Gupta et al in [15] proposed new method for detecting a target in non-stationary clutter in active sonar using dynamic time-frequency localization. In [16,17], authors try to estimate behavior of target motion by using Target Motion analysis (TMA) and Bearing Only Tracking (BOT) algorithms. Note that TMA and BOT are used after detection.…”

Sensor Network Tomography: The Revenge of the Detected

“…The strong law of large numbers ensures that, as : (18) where convergence holds with probability one. Since (19) we have…”

“…The question we pose is: Can the target exploit some knowledge of what it hears from the sensors to know where they are? The only similar example of which we are aware is our own [19], in which it is shown that in the restrictive target motion analysis (TMA) case of estimation of a straight line target trajectory given angle-only measurements from a two-leg platform path, that platform path is exactly inferable only if in addition to the track the target can "overhear" how confident (i.e., the posterior covariance) the platform is in its estimation of the track. Unlike [19] where the estimates are the target's clues to locate its threat, here we consider the range-based detections.…”