Distributed MHT and ML-PMHT Approaches to Multi-Sensor Passive Sonar Tracking

Abstract

This paper introduces a distributed multiple-hypothesis tracking (MHT) approach to passive sonar tracking with multiple sensors. Specific advances include the ability to contend with multiple emitted frequencies per target and the design of a statistically-consistent and stable sequential extended Kalman filter. We consider as well a maximum likelihood probabilistic multi-hypothesis tracker (ML-PMHT) and provide a qualitative comparison of the two methods with simulated sonar data. Directions for future work including a hybrid scheme that exploits the strengths of both tracking paradigms are provided.