Line-of-sight aware accurate collaborative localization based on joint TDoA and AoA measurements in UWB-MIMO environment

Ultra wide band (UWB) based localization has been extensively utilized in indoor and urban dense regions which enables centimeter-level localization accuracy over global positioning system (GPS). However, the prior UWB-based localization methods still lack better localization accuracy and there is increase anchor deployment cost. Multiple input multiple output (MIMO) was created to improve wireless data transfer, but its ability to leverage many antennas for spatial information has resulted in localization and tracking applications. For this reason, the researchers incorporated MIMO in the UWB to improve the performance of the localization system. Yet, there are limitations to guaranteeing communication dependability while using poor localization approaches. To address the issues in the existing works, we proposed UWB-collaborative localization (UWB-CollLoc) method. Initially, the optimal placement of anchor nodes in the UWB MIMO environment is done to mitigate the accuracy issues by static non-line of sight (NLOS) blockages. The optimal placement is done by utilizing the smart crow search optimization (SCSO) algorithm which optimally places the anchor nodes by ensuring several metrics. Before localizing unlocalized mobile orphan tags, the available line of sight (LOS) anchors and Sub-anchors in the environment are detected using lightweight attention network (LAN) by the mobile orphan tags by considering several LOS and NLOS-related metrics. After determining the LOS anchors and sub-anchors, collaborative localization is done to improve localization efficiency, reduce the new anchor deployment cost, and mitigate the errors caused by dynamic NLOS blockages. The time difference of arrival (TDoA) and Angle of Arrival (AoA)-based collaborative localization is employed to reap the complete advantages of multilateration and angulation techniques in which the already localized mobile tags act as sub-anchors for unlocalized mobile orphan tags. Finally, the correctness of the localization results is ensured by forwarding the results to the UWB Real-Time Location System (UWB-RTLS) server which validates the results using the Jaccard similarity measure. The proposed work is experimented with in the Network Simulator Version-3.26 (NS-3.26) tool and compared with existing works in terms of several validation metrics. The results show that the proposed work outperforms better than the existing works. The experimental result shows that the proposed method has number of transmitted packets (26), LOS accuracy (0.79), NLOS accuracy (0.9), and correct rate (0.905).