In order to improve the ability of the tracking loop to filter out noise , a Kalman filter estimation algorithm is proposed to track the signal in high dynamic or weak signal environment steadily. Firstly, the influence of noise on the tracking loop is analyzed. The theoretical design basis of the Kalman filter estimation algorithm is discussed, and the adaptive Kalman filter tracking loop is designed from the aspects of loop structure, model establishment and optimization design. The experimental results show that the positioning performance and accuracy are improved greatly compared with the traditional tracking loop in both static and dynamic scenes.