基于快速解卷积的稀疏时水声信道估计方法  

An Underwater Acoustic Channel Estimation Method Based on the Fast Deconvolution Method during Sparse Time-Varying

作　　者：李佩 杨益新[1] 王永威[1] LI Pei;YANG Yi-xin;WANG Yong-wei(Northwestern Polytechnical University, Xi’an 710072, China)

机构地区：[1]西北工业大学,陕西西安710072

出　　处：《山东农业大学学报（自然科学版）》2018年第3期429-432,共4页Journal of Shandong Agricultural University：Natural Science Edition

摘　　要：针对传统信道估计算法的不足之处,本文提出了一种基于快速解卷积(Fast Deconvolution Method,FDM)的稀疏时变信道估计方法。FDM算法基于?1-?2范数优化模型,结合循环矩阵特征值分解方法,将大维度矩阵运算转化为可通过FFT(Fast Fourier Transform)/IFFT(Inverse Fast Fourier Transform)快速实现的向量运算,有效降低了运算复杂度。结果表明,FDM算法具有时延超分辨能力,且具有更低的信道估计旁瓣。此外,该方法在不同的正则化参数、信噪比和不同传输距离等条件下对水声信道估计具有较好的稳健性。The Fast Deconvolution Method （FDM） for sparse time-varying underwater acoustic channel is proposed toameliorate conventional channel estimation methods. Based on .1-.2 optimization model, the method combines specialstructure of circulant matrices in order to transform large-scale matric-vector computation to vector-vector manipulationswhich are efficiently implemented by FFT/IFFT. Besides, simulation results show FDM can achieve super temporalresolution, while it has a low side-lobe. The robustness of FDM is verified through various simulation experiments underdifferent situations in terms of regularization parameters, SNR, and different transmitting range.

关 键 词：快速解卷积 水声信道 信道估计 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]