基于并列卡尔曼滤波的贝叶斯盲均衡算法实现

摘要

Research on the realization of blind Bayesian equalization using parallel Kalman filtering is carried out.Two modified measures are proposed to reduce the number of Kalman filters aiming at the high complexity of the current realization method using too much filters.One is to initialize the channel symmetrically, and the other is to renew the channel according to common correlative states.It only requires one filter to deal with the states which are symmetrical or have the same correlative states.And the number of Kalman filters in the modified method is reduced largely.Simulation result shows that the modified method has low complexity and the same performance as the typical realization method.%研究了基于并列卡尔曼滤波的贝叶斯盲均衡算法的实现方法.针对目前的实现方法中并列滤波器数量众多,实现较为复杂的状况,提出了两种减少并列滤波器改进措施--信道对称初始化和基于同相关状态的信道更新处理.它们主要根据传输向量状态的对称性和相关性将它们分成一组,每组仅采用一个滤波器来实现,使并列滤波器数最减少为原来的 1/2 M(M 为信号的调制阶数).经实验证明改进的实现方法不但大大降低了实现复杂度,而且和通常的实现方法具有相同的误码性能.