Abstract: Aiming at the problem of mobile node localization in underwater acoustic sensor network, this paper first studies the multilateral localization method based on distance measurements in detail. Then this paper presents a method of using extended Kalman filter (EKF) with the motion information to track the mobile nodes. Finally, the modified extend Kalman filter (MEKF) is proposed to improve EKF because the measurements of mobile nodes are asynchronous. The simulation study indicates that the localization accuracy of MEKF is much better than that of EKF, and EKF performs better than ML.