Non-line-of-sight identification via phase difference statistics across two-antenna elements

This study proposes and investigates the performance of a new non-line-of-sight (NLOS) identification technique for multiple antenna systems that is based on the phase difference across two antenna elements. In order to avoid any confusion in the position location process, NLOS identification techniques should be implemented in the localisation systems. A phase wrapping selection algorithm is proposed to calculate the phase difference variance across two-antenna elements. A theoretical relationship is maintained between the phase difference variance and the Rician K-factor. The proposed K estimator requires an uncorrelated phase across antenna elements. The validity of this assumption is verified via channel modelling simulations. Then, a hypothesis test on the K-factor is formed to identify NLOS situations. The prior distributions of K-factor under line-of-sight (LOS) and NLOS conditions, and the K-factor threshold that are used to distinguish LOS and NLOS situation are derived. The impact of shadowing on the performance of the proposed NLOS identification method is studied. The performance of the proposed phase difference K estimator technique is compared with the envelope-based K estimator.