论文标题:多输入多输出系统盲源分离频域新方法的研究 Research on the New Approach to MIMO Blind Source Separation in Frequency Domain 论文作者 彭帆 论文导师 戴宪华,论文学位 硕士,论文专业 信号与信息处理 论文单位 汕头大学,点击次数 866,论文页数 69页File Size2238k 2002-05-01论文网 http://www.lw23.com/lunwen_47442/ 多输入多输出系统;盲源分离;累计量;瞬时混合系统;频域盲源分离;线性卷积混合系统 MIMO system;Blind source separation;Cumulant;Instantaneous mixture system;Frequency domain separation;Linear convolution system 盲信号处理是信号处理的一种基础技术,它要解决仅仅依靠输出信号去重现整个传输系统的问题。多输入多输出系统的盲源分离,是指在对源信号和信道没有任何先验知识的情况下,从观测到的几个信号中来恢复出未知的源信号。盲源分离问题发生在很多领域,如噪声抑制,雷达和声纳处理以及语音增强等等。 在本论文中,对多输入多输出系统盲源分离方法进行了系统的研究。论文包括了以下几个部分: 1.简介 介绍“盲”的概念,并对瞬时混合盲源分离系统进行建模。 2.对瞬时混合多输入多输出系统的研究 对瞬时混合系统的盲源分离原理和技术进行了简单的回顾。 3.对线性卷积混合系统时域算法的研究 对线性卷积混合系统建模,并对时域的盲解卷方法进行研究和比较。 4.提出了一种对线性卷积混合系统进行分离的新方法 提出一种多输入多输出线性混合系统的频域盲源分离方法,进行仿真实验,以验证该算法的效果,初步结果令人满意的, Blind signal processing is a fundamental signal processing technology aimed at retrieving a system"s unknown information from its output only. MIMO blind source separation may consist of recovering unobserved sources from several observed signals without prior knowledge of sources or the channel. The blind source separation issue arises in many fields such as noise reduction, radar and sonar processing and speech enhancement.In this paper, the approaches to MIMO blind source separation are systemically studied. The contributions of this paper are followed:1 IntroductionIntroduce the concept "Blind" and present MEMO blind source separation model.2 Research on the instantaneous mixture systemReview the techniques developed for the demixing of the instantaneous mixture case.3 Research on the linear convolutive systemModel the linear convolutive system and review the deconvolution approaches in time domain.4 Introduce a new method to MIMO BSSA new approach to MIMO blind source separation in frequency domain is introduced, experimental tests are conducted,and results are satisfied.
|