【作者】 邬冬临；

【导师】 刘宏立；

【作者基本信息】 湖南大学 ， 电子科学与技术， 2010， 硕士

【摘要】 在码分多址(CDMA)系统中多个用户均占用同一时隙、同一频隙,不同的是选取的地址码不相同,在实际应用中,地址码又不可能达到理想的状态,因此在多个用户通信时,一定会产生多址干扰(MAI)。随着用户数目的不断变化,多址干扰就成为CDMA系统中最主要的干扰。多用户检测技术则是充分利用所有用户的信息来做检测,抑制多址干扰,提高系统的容量,解决了远近效应问题。在实际应用中要求多用户检测算法既能满足性能优良又能满足复杂度低,然而现有的多用户检测技术还不能在这两者之间取得平衡,因此有必要对多用户检测进行更深入细致的研究。本文首先对直接序列扩频码分多址(DS-CDMA)系统的多用户检测技术的理论与技术进行系统的介绍,然后从最原始的传统检测机到盲自适应检测机进行介绍,并对多用户检测的应用及应用前景进行探讨,对线性盲自适应多用户检测算法进行具体的比较和分析。在此基础上,本文给出了一种改进的基于压缩预测逼近子空间跟踪算法(iPASTd),以及一种基于改进的压缩预测逼近算法(iPASTd)和Kalman滤波相结合的盲自适应多用户的检测器。仿真实验结果表明,本文给出的盲自适应多用户检测器与基于最小二乘(RLS)算法的盲自适应多用户检测器和基于最小均方(LMS)算法的盲自适应多用户检测器相比较具有良好的跟踪性能,较低的复杂度,在加性高斯白噪声(AWGN)信道中取得良好的信干比,并且在高信噪比和强远近效应的情况下,本文给出的盲自适应多用户检测器能实现更快的全局收敛并且具有良好的跟踪性能。更多还原

【Abstract】 In the CDMA system multiple users are housed in the same slot and the same band gap, but select a different address code.In practice, the address code can not possibly meet ideal conditions. when multiple users communicating, it will produce multiple access interference (MAI). With the changing number of users, multiple access interference has become the most important interference in CDMA systems. Multi-user detection technology is the full use of all user information to perform detection, suppress multiple access interference and increase system capacity. In practice multi-user detection algorithm is required to meet for both excellent performance and can meet the low complexity, while the natural existing multi-user detection techniques can not achieve balance between these two. Hence the need for more multi-user detection intensive research.First the article introduce multiuser detection theory and technology in DS-CDMA systems. Then the most primitive from to the traditional blind adaptive detector testing machines are introduced in the paper. The application and the application prospect of multi-user detection will be introduced in the paper, then the paper will do a specific comparison and analysis for linear blind adaptive multiuser detection algorithm. On this basis, the paper gives an improved prediction based compression approximation subspace tracking algorithm (iPASTd), as well as an improved forecast based on improved approximation algorithm for the compression (iPASTd) and the Kalman filter combining blind adaptive multi-user detector. Simulation results show that given the blind adaptive multiuser detection algorithm based on RLS blind adaptive multiuser detector, and based on LMS algorithm for blind adaptive multiuser detector compared with a good tracking performance, lower complexity, in the AWGN channel to obtain a good signal to interference ratio, and high SNR and strong near-far situation, this paper, blind adaptive multiuser detector can achieve a faster global convergence and good tracking performance.更多还原