Utilization of Multiple-Antenna Multicarrier Systems and NLOS Mitigation for Accurate Wireless Indoor Positioning

• Published in: IEEE transactions on wireless communications 2016-10, Vol.15 (10), p.6570-6584
• Main Authors: Gaber, Abdo, Omar, Abbas
• Format: Article
• Language: English
• Subjects: Antenna arrays; Cramér-Rao lower bound; Delay effects; Delays; Direction-of-arrival estimation; DOA estimation; Estimation; NLOS mitigation; OFDM; TDOA estimation; unitary matrix pencil; Wireless communication
• ISSN: 1536-1276; 1558-2248
• DOI: 10.1109/TWC.2016.2585645

In this paper, we investigate the performance of wideband orthogonal multicarrier signals in conjunction with multiple antenna systems for wireless indoor positioning. The multiple output orthogonal frequency division multiplexing (OFDM) scheme is used to estimate the propagation time delay and the relative direction of arrival (DOA) of multipath signals. Besides the frequency diversity attributed to OFDM, the antenna array elements can be used in two scenarios: 1) for time delay estimation with spatial diversity and 2) for joint time delay and DOA estimation. To measure the effect of the characteristic parameters of wireless positioning system, the Cramér-Rao lower bound for both scenarios have been derived. In indoor environments, the mobile unit (MU) is often in a non-line-of-sight (NLOS) situation and the direct path could be completely blocked. This degrades the performance significantly. Therefore, we introduce properly weighted least square (WLS) and hybrid WLS estimators to mitigate the effect of undetected direct path channel profiles. An adaptive method is also presented to select the reference base station (BS) and extract certain weights to scale the observations based on the estimated channel profiles. Our experimental results using the emerging IEEE 802.11ac standard reveal that the hybrid time difference of arrival (TDOA) and DOA outperform in non-perfectly synchronized BSs and NLOS environments. However, the achieved accuracy is not improved beyond that obtained by using TDOA with spatial diversity if the BSs are perfectly synchronized. The NLOS mitigation method has been tested using different obstacles with different positions between the MU and the BS. The achieved performance is nearly similar to that of using LOS BSs.