Time difference localization algorithm based on modified unscented Kalman filter

For the less precision or divergence problem of TDOA localization algorithm in the condition of NLOS propagation. In this paper, improvements are introduced to enhance UKF: judging the type of NLOS error and using the total-deflection method can modify Kalman gain to mitigate NLOS error effectively and realize LOS reconstruction, and estimate covariance is not modified to avoid divergence. Finally Simulation results demonstrate that the proposed method can restrain the influence of mutation, but also mitigate NLOS error highly effectively and improve position precision greatly.