Multiple Hypothesis Data Association for Multipath-Assisted Positioning

Abstract

Global navigation satellite system denied scenarios such as urban canyons or indoors cause a need for alternative precise localization systems. Our approach uses terrestrial signals of opportunity in a multipath-assisted positioning scheme. In multipath-assisted positioning, each multipath component arriving at a receiver is treated as a line-of-sight signal from a virtual transmitter. While the locations of the virtual transmitters are unknown, they can be estimated simultaneously to the user position using a simultaneous localization and mapping (SLAM) approach. An essential feature of SLAM is data association. This paper addresses the data association problem in multipath-assisted positioning, i.e., the identification of correspondences among physical or virtual transmitters. If a user recognizes a previously observed transmitter, it can correct its own position estimate. We generalize a previous version of our multiple hypothesis tracking scheme for data association in multipath-assisted positioning and show by means of simulations how data association improves the positioning accuracy.