一种可变遗忘因子的卡尔曼滤波信道跟踪

A variable forgetting factor of Kalman filter channel tracking

在快衰落信道下,时变特性会使导频时段与数据时段的信道冲激响应存在误差,所以需要引入信道跟踪。实践证明,系统建模与实际模型存在一定偏差时,标准卡尔曼滤波的鲁棒性较差。因此,文中提出了一种可变遗忘因子的卡尔曼滤波算法。仿真表明,与标准卡尔曼滤波相比,可变遗忘因子的卡尔曼滤波信道跟踪具有更好的鲁棒性。

In fast fading channel, because of the time-varying characteristics, there is a deviation of the channel impulse between the pilot period and data slots, so it should be introduced into the channel tracking module. Practice has proved that the system modeling and the actual model there is a certain deviation which makes the standard Kalman filter robustness poor. Therefore, this paper proposes a variable forgetting factor Kalman filter algorithm. The simulation resuhs show that, compared with the standard Kalman filter algorithm, the variable forgetting factor of Kalman filter channel tracking is more robustness.