A localization technique for LOS and NLOS scenario

Various localization techniques are described in the literature, but are in many cases, accurate only under LOS conditions. To cope the bias effect of reflections when the mobile transmitter Tx moves in an indoor environment, we propose a method using at least two fixed receivers Rx1 and Rx2 and based on the extraction, for each link pair Tx-Rx1 and Tx-Rx2, of the characteristics of the first two paths in terms of direction of arrival (DOA), direction of departure (DOD) and time of arrival (TOA). These two paths can be either the direct path and the first reflected path or the first two reflected paths, for LOS and NLOS environments, respectively. To determine the paths characteristics, uniform rectangular arrays (URA) are used at Tx, Rx1 and Rx2, the parameters estimation being performed through high-resolution techniques. Since on one hand, the orientation of the mobile array Tx is unknown, and, on the other hand, only relative values of TOA and DOD are measured to eliminate synchronisation between Tx and Rx, i.e. the angle between two rays departing from Tx and the time interval between rays arriving at the receiving points. Additionally, the location of the mobile and the reflection points are unknown. This leads to a set of non-linear equations with ten unknown values, which is solved numerically. In this paper, the principle of this localization technique is described and preliminary results are presented for a simulated environment.