一种水声时变信道低计算量自适应均衡算法

对于时变水声信道，迭代步长的取值对均衡算法的收敛特性影响较大，步长的自适应参数选取解决了上述问题，但随之而来的是计算量的显著增加。提出了一种自适应均衡算法，该算法根据均方误差切换准则在两种算法之间进行切换。对不同水声信道进行了大量的计算机仿真，结果验证了本算法的有效性。该算法解决了时变信道迭代步长选取困难的问题，同时将均衡计算量降低了30％，更适合快速变化水声信道下的实时通信。

Low Computational Complexity Adaptive Equalization Algorithm for Time-variety Underwater Acoustic Channels

The performance of equalization algorithm on different channels is sensitive to the choice of step-size or forgetting factor. To overcome this problem, self-optimized algorithm may be used, but this results in increased complexity. A new equalization algorithm was proposed which had an algorithm transition under a mean square error (MSE) rule. Simulation results on different underwater acoustic channels show that, unsuitable step-size causes algorithm diverge, while the approach has robust convergence performance. At the same time it can decrease 30% computation complexity compared with FOLMS algorithm. This algorithm is particularly suitable for real-time implementation under time-variety underwater acoustic channels for its low computational load and self-adaptive behavior.