【作者】 黄海洋；

【导师】 李少谦；

【作者基本信息】 电子科技大学 ， 通信与信息系统， 2009， 博士

【摘要】 多天线预编码是新一代无线通信的核心技术之一。基于多天线的开环MIMO技术能够利用空间自由度获得复用和分集增益,以此提高无线通信的频谱效率和功率效率。多天线预编码在此基础上利用系统后端传递到前端的辅助信息对待发送的源信号进行针对性预处理,则可以进一步提高系统综合性能,并且为降低系统接收机的设计复杂度提供了可能。本论文围绕着多天线预编码技术领域中的核心热点问题,在物理层层面上,从信号传输角度出发,对点对点多天线系统的线性预编码设计,多天线预编码系统中的反馈数据压缩策略,多用户多天线系统的优化预编码设计,以及预编码在MIMO-OFDM系统中对峰均功率比的抑制等问题进行了深入的创新性研究。这些研究力求探索信号传输本质,发掘系统性能潜力。鉴于工程可行性要求,性能和复杂度的折中优化亦是本论文的考虑方面。全文主要内容归纳如下。第一章介绍了多天线预编码技术的研究背景,包括和本文工作密切相关的技术定义,分类,历史以及该领域研究的特点。特别总结了部分经典技术以备后文参考。第二章提出两种创新的点对点多天线线性预编码优化设计。第一种是基于螺纹代数空时(TAST)码的线性预编码方案。针对该空时码的内在结构特点,优化了预编码码字选择方法,提出一种称之为奇异值之积选择准则(PS-SC)。数学分析和仿真结果均表明其性能和复杂度的折中比既有方案更理想。第二种是基于正交空时分组码(OSTBC)的非酉的线性预编码设计。根据OSTBC的独特结构,在信号总发射功率不变且不增加系统实时处理复杂度的前提下,利用非酉结构的预编码设计使得理论优化和对应的码本量化的系统综合性能都比传统方案有显著提高。同时,给出并论证了该方法保证系统满分集时对码本大小的约束条件。在此基础上,针对天线空间相关的情况,提出一种改善性能的融合“注水”的预编码码本生成解决方案。多载波技术(比如OFDM)的引入使得预编码的反馈数据量可能随着载波数线性增加,而在实际中未必能够容忍。第三章针对该情况,从预编码码字跟踪的策略角度,根据系统性能界的优化,提出一种新的测度。在保持较低反馈信息量的前提下,该测度能够进一步改善系统性能,并且同时适用于频域相关和时域相关信道。由于该测度方案算法复杂度较低,所以具有工程实用价值。利用发射空间自由度进行用户间数据分离的多用户多天线预编码技术是当前的一个热门研究方向。第四章提出了一种改进的基于信漏噪比(SLNR)的预编码方案。它包括分别侧重不同信噪比(SNR)区域优化的两种迭代算法。该方案深入发掘系统设计中的预编码矩阵和接收合并矩阵之间的内在联系,利用新提出的迭代算法解决其本质上存在的嵌套关系完成求解,从而实现性能优化。数字仿真从平均误符号率(SER)与系统和速率(sum rate)两个角度进行验证。结果表明,新方案相较于原始基于SLNR的方案以及其它经典方案具有综合性能优势。而且,由于迭代算法收敛迅速,因此有利于工程实现。第五章讨论了利用预编码来解决基于空时码的MIMO-OFDM系统的PAPR抑制问题。通过利用频率自由度进行信号变换,提出一种跨频率置换和求反(CFPI)的预编码技术。分析表明其对PAPR抑制效果明显。另外,由于该技术对空时和空频结构的MIMO-OFDM系统均兼容,因此相比既有方案具有更广泛的适用性。最后,第六章对全文进行总结,并根据目前的研究工作对未来研究方向予以展望。更多还原

【Abstract】 Multi-antenna precoding is one of the key technologies of next generation wireless communications. Employing multiple antennas, the open-loop MIMO can provide multiplexing and diversity gains by making use of the additional degrees of spatial freedom, and improve the spectral and power efficiency of wireless communications. Furthermore, if the source signal is pre-processed according to the auxiliary information, which is conveyed from the backend to the frontend of system, multi-antenna precoding can improve the overall performance with probably lower receiver complexity.In this dissertation, several hotspots in the multi-antenna precoding research field are investigated from signal transmission aspect of physical layer, including the linear precoding design in point-to-point multi-antenna systems, feedback data compression strategy of multi-antenna precoding systems, optimal precoding design of multi-user multi-antenna systems, and using precoding to suppress PAPR of MIMO-OFDM systems. These researches strive to explore the essence of signal transmission, and release the potential of system performance. In view of the practical feasibility, the proper tradeoff between performance and complexity is another consideration. The main content is summarized as follows.Chapter 1 introduces the research background of multi-antenna precoding, including the technique definitions, classification, history, and specific character of the research field, which are relevant to the innovative work of this dissertation.Two optimal linear precoding designs of point-to-point multi-antenna systems are proposed in Chapter 2.The first is the linear precoding design for TAST codes. Especially, an optimal precoding codeword selection method named PS-SC is proposed according to the intrinsic feature of TAST codes. The mathematical analysis and simulations indicate that it achieves a better tradeoff between performance and complexity than the case of previous designs. The second is the non-unitary linear precoding design for OSTBCs. Without increasing the real-time processing complexity, the overall system performance of theoretical optimization and codebook quantization case are both improved compared with the conventional design by using the non-unitary precoding structure in the sense of restriction on the total power of transmit signal. Moreover, the constraint on codebook size to guarantee full diversity order is given and proved. Based on this, to combat the antenna spatial correlations, a precoding codebook generating solution mixed with "water-filling" is proposed, which has increased performance improvement.The introduction of multi-carrier techniques (such as OFDM) increases feedback data of precoding, which may be proportional to the subcarrier number. However, this is probably not supported in practice. Chapter 3 deals with this problem from the precoding codeword tracking strategy point of view. As this, a new metric is proposed via system performance bound optimization. At the cost of low amount of feedback data, this metric improves system performance, which can be applicable to both frequency and temporal correlations of channels. For the lower complexity, it can be easily applied to practical systems.The multi-user multi-antenna precoding technique, which uses the additional degrees of transmit spatial freedom to decouple different user data, is a hot research field recently. An improved precoding method based on SLNR is proposed in Chapter 4, including two iterative algorithms whose optimizations emphasize on the different SNR regions. This method digs out the inherent relationship between the precoding matrices and receive-combining matrices, and achieves the optimized solutions by the iterative algorithms to solve the nesting problem.The advantages are verified from both the average SER and the system sum rate aspects in the numerical simulations. It is observed that the method is superior to the original design based on SLNR and many other classical methods. In addition, with the rapid convergent speed, the proposed iterative algorithms are favorable for practice.Chapter 5 discusses using precoding to suppress PAPR of MIMO-OFDM systems based on space-time coding. A cross-frequency permutation and inversion (CFPI) precoding technique is proposed, which transforms signal via additional degrees of frequency freedom. The analysis shows that the PAPR suppression is solid. Especially, CFPI has more extensive applicability compared with the previous techniques for it is compatible with either space-time or space-frequency MIMO-OFDM systems.Finally, the contributions in this dissertation are summarized in Chapter 6. The suggestions for future research are also presented.更多还原