干扰协作多基站宽带无线网络容量和干扰管理研究

Since frequency spectrum and energy are both rare resources,green wireless transmission techniques with high spectrum efficiency and high energy efficiency show important values of theory and bring large economic profit. This project focuses on the research about capacity and downlink transmission in interference cooperative multi-BS wideband green wireless networks with high spectrum efficiency and energy efficiency: the complicated channel model considering the effects of antenna polarization, shadow-fading, and interference will be established and the closed-form expressions for capacity, capacity bounds, and degrees of freedom (DoF) will be derived, these will provide an information theory basis for link transmission; wideband interference alignment (AI)schemes will be constructed to achieve the DoF of network and to realize the effective interference management; an optimal wideband power allocation algorithm is designed based on AI,game theory, and convex optimization, then an alternating iterative structure between PA and AI is developed to minimize the impact of interference, to approach the sum capacity , and to fully exploit the multiuser diversity and PA gains in multi-cell wireless networks; Finally, to achieve a high-speed and high-performance limited feedback transmission, and to reduce the effect of interference, two high-efficiency rotating differential codebooks based on vector quantization and differential manifold are designed for power allocation and precoding matrix of AI at transmitter, respectively. Our project covers the study of channel capacity, link transmission, and key patents of multi-cell MU-MIMO systems, and will offer a theory support, some key techniques, and a novel solution for the future wireless communications of our nation based on multi-cell cooperation.

频谱和能源均是稀缺资源，高频谱效率和高能效的绿色无线传输技术具有重要的理论价值和可观的经济效益。本项目研究绿色高频谱效率且高能效的干扰协作多基站宽带无线网络容量与下行链路传输技术: 建立考虑天线极化、阴影衰落以及干扰的复杂信道模型并推导对应的信道容量、容量界和自由度的解析公式，为链路传输提供信息论基础；构造可达网络自由度的宽带干扰对齐方案，实现有效的干扰管理；给出基于干扰对齐、博弈论和凸优化的最优宽带功率分配算法，然后发展基于干扰对齐和功率分配的交替迭代结构, 最小化干扰影响,逼近容量和，充分挖掘多用户分集和功率维增益；分别针对功率分配向量和干扰对齐预编码矩阵，设计基于向量量化和微分流形的高效旋转差分独立双码本, 实现有限反馈的高速率与高性能传输，同时抑制干扰。本项目涉及多蜂窝网络容量、链路传输与核心专利，为我国新一代基于多蜂窝协作的宽带无线通信网络提供理论支撑、关键技术和全新的解决方案。

考虑大尺度阴影衰落、基站发射天线相关性和用户接收天线相关性前提条件下， 建立了多蜂窝多用户MIMO 系统的复合信道模型；设计并提出了一种交替迭代结构, 该结构整合准最大化和速率预编码与基于最大化和速率准则的低复杂度交替网格搜索功率分配方案， 形成最大化和速率与低复杂度双层预编码与功率迭代结构。针对在实际的多用户多基站宽带MIMO-OFDM 系统中，发射端获得的信道状态信息（Channel state information，CSI）会受到时延、信道估计方法、导频图样、导频符号发射功率及数目等因素的影响，从而产生信道估计误差的问题，本项目组提出了一种自适应稳健的基于最大化信泄噪比的预编码算法，该算法将CSI误差模型系数建模为时延与信道训练网格处等效信噪比的函数，通过估计多普勒扩展及信噪比来估计CSI误差模型系数。为了解决时变宽带多用户MU-MIMO-OFDM系统整体预编码复杂度高问题，提出了一种低复杂度的基于最大化信泄噪比的分组(Grouped-subcarrier maximum signal-to-leakage-and-noise ratio，GS-Max-SLNR）预编码方案。通过理论分析和仿真结果表明：所提双层交替迭代结构和速率性能逼近和速率上界；所提自适应稳健预编码算法能够自适应地实时跟踪CSI的变化，具有良好的稳健性；同全局Max-SLNR预编码相比，所提出的GS-Max-SLNR预编码算法的复杂度会远远低于前者，能实现复杂度与性能的良好折衷。 本项目组所提出的信道模型，预编码与功率分配算法等可为未来多基站多用户系统奠定基础理论和链路传输方法提供有益补充和新途径，可用于新一代多蜂窝协作的宽带无线通信网络如下一代无线局域网与移动互联网络等场景。