【作者】 李姜兵；

【导师】 邓平；

【作者基本信息】 西南交通大学 ， 通信与信息系统， 2013， 硕士

【摘要】 为了适应移动互联网的快速发展,及应对IEEE联盟Wimax技术的挑战,3GPP在发布LTE Rel-8后,启动了LTE-A的标准化工作,引入了协同多点传输技术(Coordinated Multi-Point Transmission/Receiption, CoMP),旨在提高小区的平均频谱效率和小区边缘的频谱效率。协同多点传输分为协同调度和联合处理,本文主要研究联合处理。通过多个基站之间共享用户的数据信息和控制信息,相邻的小区的可以参与协作,同时向参与协同传输的用户发送数据,可以有效的减少邻区之间的干扰,从而提升用户的接收信干噪比。进而提升小区边缘用户的性能。协同多点传输技术的实现,需要面对许多问题。例如如何设计良好的预编码方案,以有效的消除多用户之间的干扰,怎样在有限反馈下取得性能和系统负担的折中等,本文主要针对这两个问题展开研究。首先,针对CoMP中联合传输预编码问题,本文重点研究了BD算法,SLNR最大化算法在完美信道状态下和有限反馈的情况下的性能。针对SLNR最大化问题进一步研究了其改进的算法,其中CISVD算法以SLNR最大化为起点,进行迭代以最大化用户的和速率；另外一种改进算法也是以SLNR最大化为起点,将预编码矩阵和接收矩阵嵌套迭代。然后,在有限反馈的情况下,研究了多用户MIMO中自适应反馈和差分反馈,研究了反馈不同形式的协方差矩阵对系统性能的影响。实验结果表明这两种方案都较Rel-8中的固定码本反馈有较好的性能提升。针对现有的CoMP的反馈框架,研究了适合相干联合处理的单个小区反馈方式,选取三个bit量化小区间的相对相位信息取得了性能和反馈负担的折中。本文将自适应反馈运用到CoMP中,相对于传统的非自适应反馈,有一定的性能提升。最后,本文研究了一种新的量化bit数自适应分配方案,将上行的反馈资源分为两部分,一部分用于CoMP用户进行多小区传输,用户的协作集的反馈bit数进行自适应分配,余下的反馈bit数对多用户MIMO用户进行两步自适应反馈。仿真结果表明,此方案不仅能提高小区边缘的频谱效率,也能取得多用户分集增益。更多还原

【Abstract】 With the rapdly development of mobile internet,3GPP have initiated the standardization of LTE-Advanced after release the first version of LTE, to cope with the challenge of Wimax, Coordinated Multi-Point Transmission and reception was involved in LTE-Advanced to improve the cell average and cell edge spectrum efficiency. CoMP was classified into Joint Process and Coordinate Schedule. This thesis foucs on Joint Process, Multiple basestations in the cooperative set transmit the data to UE simultaneously at the cost of the share of data and control information among basestations to reduce the inter cell interference, consequently the users spectrum efficiency in the cell edge was improved.Couple of issues need to be solved before CoMP was implemented.,such as how to design a fine precoding matrix to cancel the multi-user interference;how to made tradeoff between feedback overhead and system performance. These two issues were addressed in this thesis.Fistly, when it comes to the Joint Process Precoding, the performance of BD and SLNR in both perfect CSI and limited feedback was firstly detailed. And two improved algorithms with respected to SLNR were researched, which both started from SLNR. CISVD was based on maximize sum rate and another one was based on receiver precoding matrix nesting iteration.Secondly, in limited feedback case, some enhanced feedback schemes named adaptive feedback and differential feedback were researched, the impact of different type of convariance matrix was evaluated. Adaptive and differential feedback both outperformed raditional feedback scheme. According to the3GPP contribution, an appropriate feedback scheme was researched. Simulation results demonstrated that3bits to quantized inter-CSI-RS information performed well in feedback overhead and system performance improvement tradeoff. Then a new adaptive feedback scheme was proposed for CoMP JP and won performance improvement.Finally a new adaptive feedback allocate scheme was proposed, the feedback bits were classified into two parts. Some of the bits are occupied by CoMP users, which were adaptively allocated for different cells according to the channel quality. And the rest of the feedback bits were used by single cell MU-MIMO users for two stage adaptive feedback. And simulation results demonstrated the proposed scheme perform well in cell edge and gain multi-user diversity.更多还原