In a radar target tracking system in a complex environment, the noise pollution distribution problem is caused by external interference. In order to ensure the reliable implementation of the system, an innovation-based adaptive EKF radar target tracking algorithm (IAEKF) is proposed. By establishing the statistical characteristics of system innovation covariance is directly introduced into the filter gain matrix calculation. The algorithm improves the adaptability of the algorithm without increasing the computational cost. Simulation test show that the IAEKF algorithm has higher tracking accuracy than the EKF algorithm in the case of radar measurement noise pollution distribution. The algorithm is feasible and effective, and has certain engineering research value.