Discerning Direct and Indirect Paths: Principle and Application in Passive Target Positioning Systems

摘要

Passive target localization is an important prerequisite for applications ranging from security to logistics to emergency response. The localization system estimates the location of the target by utilizing runtime and direction of the signal (multipath component, MPC) directly reflected by the target. Indirect paths (IPs), i.e., MPCs that interact with environmental objects as well as the target, need to be discerned from the direct paths; failing to do so brings bias to location estimate, and is a major source of localization error. This paper proposes a novel scheme for the detection of IPs in the presence of noise, jointly employing the time of arrival (TOA), direction of arrival (DOA) and direction of departure (DOD) estimates. The decision statistic of this scheme is the distance between the intersection of DOA and DOD directions and the TOA ellipse. Under the assumption of Gaussian errors of TOA, DOA and DOD, the scheme is implemented to maximize the probability of detection given a fixed probability of false alarm. The paper derives closed-form equations for the decision criterion and the resulting performance. Simulation results are given to support the validity of our scheme.