Cooperative multipath-assisted indoor navigation and tracking (Co-MINT) using UWB signals

In multipath-assisted indoor navigation and tracking (MINT), explicit use is made of multipath propagation in the ultra-wideband channel. With the help of floorplan information, localization is possible with only one reference node. In this work, we introduce MINT among cooperating users in order to omit the need of any known reference nodes. The received signal is modeled as a combination of deterministic multipath components, diffuse multipath represented by a random process and AWGN. In a mixed line-of-sight (LOS)/non-LOS (NLOS) indoor scenario, we show how information from mono-static and bi-static measurements of cooperating users can be merged for localization and tracking. The problem is formulated with a factor graph and solved via belief propagation on the factor graph. Simulation results show that localization and tracking of a mobile agent is possible: (i) independent of LOS or NLOS, (ii) without the need for further infrastructure.