Robust Single-point Localization Technique Using Downlink TDOA-AOA Fusion

In indoor environments, conventional Global Navigation Satellite System (GNSS) is difficult to provide accurate localization due to signal interference. In this paper, we propose a robust single-point localization technique using downlink TDOAAOA fusion for self-localization systems, which can be applied to hierarchical self-organized wireless sensor network (WSN) to provide tags with localization that does not consume communication bandwidth. The positioning algorithm proposed in this paper utilizes a planar antenna array to receive the broadcast signal from base stations (BS), and constructs a nonlinear least-squares optimization problem to achieve single-point localization by fusing the time difference of arrival (TDOA) and angle of arrival (AOA) measurements. Aiming at the non-line-of-sight (NLOS) problem in real indoor scenarios, we establish an outlier isolation strategy (OIS) based on the geometrical constraints between the tag and BS, which is combined with M-estimation to realize the robust localization algorithm. Real-world experiments conducted in an underground parking lot verify the feasibility and localization accuracy of the proposed technique.