【作者】 郭玉华；

【导师】 艾文宝；

【作者基本信息】 北京邮电大学 ， 运筹学与控制论， 2013， 硕士

【摘要】 多输入多输出(MIMO, Multiple-Input Multiple-Output)系统是近几年飞速发展的无线通信系统核心技术之一,不仅可以用来提高信道容量,还可以提高信号的信噪比(SNR, Signal And Noise Ratio)。在IEEE802.11n (Wi-Fi),3GPP LTE, WiMAX, HSPA+等技术标准中已经融合了MIMO系统。MIMO系统主要有三种技术,分别是预编码技术,空间复用技术和分集编码技术。其中预编码技术可以自适应的调整用户的预编码矩阵,进而抑制信道衰落、降低系统误码率、扩大系统容量。相比传统的单用户MIMO预编码技术,多用户MIMO预编码技术需要处理多个并行的数据流,并且还要确保能够消除共信道干扰(CCI,Co-Channel Interference)。多用户MIMO预编码技术可以分为基于信道分解的预编码和最优化预编码。本论文主要对下行链路最优化预编码以及其相关技术进行了研究。下行链路最优化预编码模型是一个带一个二次不等式约束的分式优化问题。目前通常的办法是直接将分子看做目标函数(忽视分母)进行求解,毫无疑问,这样计算出来的结果误差很大。本文通过把分母从分式中分离出来作为约束条件,获得了与原问题等价的带两个约束的二次优化问题,接着使用半正定松弛方法和矩阵秩一分解技术对该等价问题精确求解。仿真结果表明本文所求结果的误差远远小于目前使用方法造成的误差,初步显示本文所提预编码技术效果显著。更多还原

【Abstract】 Multiple-Input Multiple-Output (MIMO) system is one of the important technologies in wireless communication system. It will not only improve the channel capacity, but also rise the Ratio of Signal and Noise (SNR). The IEEE802.11n(Wi-Fi),3GPP LTE, WiMAX and HSPA+have contained MIMO system. In general, MIMO system has three technologies, that is Pre-coding technology, Spatial multiplexing technology and Diversity coding technology. Pre-coding technology can adjust pre-coding matrices of users adaptively to suppress channel fading, to reduce system’s Bit Error Ratio (BER) and to improve system capacity.Different from the traditional one-single-user MIMO pre-coding technology, Multi-User MIMO (MU-MIMO) pre-coding technology has to process multiple parallel data stream, and to ensure removement of Co-Channel Interference (CCI). MU-MIMO pre-coding technology could be divided into pre-coding based on channel decomposition and optimal pre-coding.In this paper, we have discussed downlink optimal pre-coding technology. Downlink optimal pre-coding model is a fractional optimization problem with one quadratic inequality constraint. As we know, the current popular approach for solving the downlink optimal pre-coding model is to optimize only the numerator of the fractional objective function (that is, ignore its denominator), which, of course, causes great errors. In this paper, we manage to transform the original problem into a quadratic optimization problem with two constraints equivalently, by moving the denominator to a constraint. Then we solve the new problem accurately by semi-definite programming relaxation method and rank-one matrix-decomposition technique. Simulation results show that the errors caused by our algorithm are far less than those of the current approach, that is the new approach is effective.更多还原