NLOS Links Detection Algorithm for Improved 5G NR Indoor Positioning

In this paper we derive a signal processing algorithm for NLOS links detection and propose a reliability function that can be used in the positioning equations to perform correct weighting of the measurements. It is shown that the measurements may suffer from the NLOS propagation phenomenon, specifically in indoor environments, that causes an excess delay relative to the LOS transmission time. The proposed algorithm performs a binary hypothesis testing between the LOS and NLOS channel hypothesis to identify the link type. It uses an average NLOS channel power, estimated per subcarrier basis in the frequency domain to perform links classification. The provided simulation results demonstrate the validity of presented analytical formulation and the feasibility to achieve a submeter level positioning accuracy in application to the Industrial Internet of Things (I-IoT) use case.