Robust tracking and geolocation for wireless networks in NLOS environments

We address the problem of robust tracking and geolocation using time of arrival estimates in wireless networks. Especially in urban or indoor environments and hilly terrains, these estimates are often contaminated by interference due to non-line-of-sight (NLOS) propagation. Standard techniques such as least-squares are inadequate as they lead to erroneous position estimates. We propose robust methods for tracking and geolocation based on a semi-parametric approach that does not require specification of the noise density. Unlike conventional, minimax based, robust techniques, we show that our proposed techniques are more robust as they adapt automatically to the interfering environment. Specifically, we propose a robust extended Kalman filter for tracking a mobile terminal based on robust semi-parametric estimators. Numerical studies for different network scenarios illustrate a substantial gain in performance compared to standard robust competitors.