Improving Maps of Physical and Virtual Radio Transmitters

Kurzfassung

In multipath assisted positioning, each multipath component arriving at a receiver is regarded as being sent by a virtual transmitter in a line-of-sight condition. These virtual transmitters can be used for localizing a user. In the general case, their locations are unknown. With simultaneous localization and mapping (SLAM), the locations of physical and virtual transmitters can be estimated jointly with the location of a user. However, the convergence time can be high. In scenarios such as malls or museums, multiple users travel through the same scenario on different trajectories. Different users can cooperate by exchanging maps of physical and virtual transmitters in order to decrease the convergence time and increase the positioning performance. Though, each user is in its own local coordinate system. When a map of transmitters is handed from one user to another, the relative rotation and translation of the coordinate systems of the two users need to be estimated based on the transmitters in the two users’ maps. Within this paper, we propose an algorithm to merge and update an obtained transmitter map with the transmitters estimated by the current user. We evaluate our proposed method by simulations in an indoor mall. In a crowdsourcing based approach, maps are handed from one user to another, updated, and handed to the next user. Our evaluations show that exchanging transmitter maps improves the positioning performance, although each user is in its own local coordinate system not knowing the parameters relating its own coordinate system to the ones of obtained transmitter maps.