A combined approach for NLOS mitigation in cellular positioning with TOA measurements

In this paper, we propose a method to mitigate the effects of non line of sight errors in range-based localization of mobile phones. The TOA/range measurement errors are dynamically tracked with a Kalman filter, which are then combined through a constrained weighted least squares (CWLS) algorithm to locate a mobile station with respect to three or more base stations. NLOS situations, which cause a large error in TOA readings, are initially detected and handled with a highly biased Kalman filter. The use of CWLS further mitigates NLOS errors. Simulation results clearly demonstrate the improvement offered by this combined approach.