基于AR模型的水声信道盲估计算法

An Underwater Acoustic Blind Channel Estimation Algorithm Based on AR Models

给出了基于AR模型的水下通信系统模型,利用基于输出序列自相关阵的Levinson-Durbin递推算法,在不需要训练序列的条件下估计得到信道传递函数,并恢复出原信号。与基于高阶统计特性方法相比,该算法运算复杂度低,估计速度快,更适合高速数据传输信道的估计。通过对水声信道参数盲估计的计算机仿真,验证了该算法具有较高的精度。

This paper proposes an AR model-based underwater acoustic communication system. The Levinson-Durbin recursion algorithm, which is based on autocorrelation matrix of output sequence in the underwater communication system,can be applied to estimate the channel function and recover the original communication signal. Compared with high order statistics approaches this method has less complexity and is more suitable for high speed communication channel estimation. Simulation results of underwater acoustic channel estimation indicate that the algorithm has a high estimation accuracy.