多通道盲反卷积算法综述

A Review of Multichannel Blind Deconvolution Algorithm

本文阐述了多通道盲反卷积(Multichannel Blind Deconvolution-MBD)的发展历程、基本模型及假设、数学原理以及通用求解过程,讨论了目前MBD几个研究方向的发展现状与面临的问题。在分析了MBD研究进展的基础上,从时域、频域、时频域、子空间、子带技术及其他算法六个方面分类综述和比较了MBD典型算法的特点与方法思路,说明了各类算法的优缺点以及主要问题。最后指出了目前多通道盲反卷积算法研究存在的不足,并提出了MBD未来理论和应用研究中可继续深入的开放课题。

The history, basic model, hypothesis, mathematical principle, and general solutions of multichannel blind deconvolution （MBD） are described in this review. One dimension signals such as voice, communication and radar signals under the condition of MBD are the main object of study. Then, the state of the art and the challenge problems in MBD research are discussed. On the basis of the latest process, the characteristic and methods of typical algorithms are summarized from the six aspects which are time domain, frequency domain, time-frequency domain, subspace, sub-band technology and other fields. The major defects, advantages and disadvantages of these algorithms are proposed. Finally, the existing problems on the research of MBD are generalized, and the open areas of theoretic and applied research about the noisy, underdetermined, nonlinear, non-stationary, and standardized problems in combination with MBD are brought forward.