基于ICA的扩频序列盲估计算法研究

发布时间：2018-04-05 10:31

本文选题：扩频序列盲估计　切入点：独立分量分析　出处：《东北大学》2014年硕士论文

【摘要】：直接序列码分多址(Direct Sequence -Code Division Multiple Access, DS-CDMA)是一种使用直接序列扩频技术进行码分多址的常用通信系统,在非合作情况中,直扩信号的扩频序列估计方法对实现盲解扩以恢复出用户信息至关重要,因此开展DS-CDMA系统的扩频序列盲估计研究意义重大。独立分量分析(Independent Component Analysis, ICA)作为被广泛应用的盲源分离方法,用于扩频序列估计问题时,可以充分利用各用户扩频序列相互独立的特点,具有实现复杂度低,对先验知识要求较少等特点,目前已成为研究热点。本文围绕基于独立分量分析的扩频序列盲估计算法展开研究。主要工作如下：首先介绍了DS-CDMA通信系统的相关知识内容,并分析得到了DS-CDMA扩频序列盲估计的模型。然后对独立分量分析的基本理论进行了阐述,列举了ICA中常用的目标函数和寻优算法。其次,为了得到更为准确的扩频序列估计结果,在充分考虑扩频序列估计问题数据特点的基础上,对传统FastICA算法进行优化,分别从白化角度和寻优角度出发,结合核熵分量分析(Kernel Entropy Component Analysis, KECA)方法和蝙蝠(BAT)算法,提出了KECA-ICA算法和BAT-ICA算法。仿真结果表明,和FastICA算法相比,两种优化算法的扩频序列估计结果更为准确。进一步地,本文综合此两种优化算法优点,将KECA-BAT-ICA算法应用到扩频序列盲估计中,仿真结果证明了它的优越性。再次,针对用户数未知时的扩频序列盲估计问题,仍以ICA盲源分离为前提,提出一种基于信号归类的扩频序列估计算法。与传统方法相比,该方法不仅省略了复杂的源信号数估计步骤,并且避免了由于源信号数估计错误导致的扩频序列恢复失败的可能。仿真证实了新方法用于扩频序列估计时的有效性和相对于传统方法的优越性。最后,对本文的研究工作进行了总结,并对未来的研究方向进行了展望。
[Abstract]:Direct Sequence-Code Division Multiple access (DS-CDMA) is a common communication system using direct sequence spread spectrum technology for code division multiple access (CDMA).The spread spectrum sequence estimation method of direct spread spectrum signal is very important to realize blind despreading to recover user information, so it is very important to study the spread spectrum sequence blind estimation in DS-CDMA system.As a widely used blind source separation method, Independent Component Analysis (ICA) can make full use of the independent characteristics of each user's spread spectrum sequence when it is used in the spread spectrum sequence estimation problem.At present, it has become a research hotspot for the characteristics of less prior knowledge.This paper focuses on the blind estimation algorithm of spread spectrum sequences based on independent component analysis (ICA).The main work is as follows: firstly, the related knowledge of DS-CDMA communication system is introduced, and the blind estimation model of DS-CDMA spread spectrum sequence is obtained.Then the basic theory of independent component analysis (ICA) is expounded, and the objective functions and optimization algorithms commonly used in ICA are listed.Secondly, in order to obtain more accurate results of spread spectrum sequence estimation, the traditional FastICA algorithm is optimized on the basis of considering the characteristics of the data of the spread spectrum sequence estimation problem, respectively from the angle of whitening and optimization.Combining kernel entropy component analysis with Kernel Entropy Component Analysis (KECA) method and bat BATalgorithm, the KECA-ICA algorithm and BAT-ICA algorithm are proposed.The simulation results show that compared with the FastICA algorithm, the spread spectrum sequence estimation results of the two optimization algorithms are more accurate.Furthermore, this paper synthesizes the advantages of these two optimization algorithms, and applies the KECA-BAT-ICA algorithm to the blind estimation of spread spectrum sequences. The simulation results prove its superiority.Thirdly, for the blind estimation of spread spectrum sequences with unknown number of users, a spread spectrum sequence estimation algorithm based on signal categorization is proposed based on ICA blind source separation.Compared with the traditional method, the proposed method not only omits the complicated steps of source signal number estimation, but also avoids the possibility of the spread spectrum sequence recovery failure caused by the error of source signal number estimation.The simulation results show that the new method is more effective than the traditional method in the estimation of spread spectrum sequences.Finally, the research work of this paper is summarized, and the future research direction is prospected.
【学位授予单位】：东北大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN914.42

【相似文献】