时变信道下运用快速递归提取实现MIMO通信系统盲均衡的方法研究

Due to the advantages of high mobility and large system capacity, the high/medium speed mobile Multiple-Input Multiple-Output (MIMO) communication systems have attracted much attention. Blind equalization ensures the high spectrum efficiency while compensating the channel distortion. However, the static channel would evolve into dynamic channel when the system moves at a high/medium speed, and the blind equalization methods for time-varying MIMO systems exist the problems of local convergence, slow convergence and high computational complexity. This project intends to utilize the fast recursive extraction method to study the blind equalization for time-varying MIMO systems and solve these problems. Firstly, combined with the space-time diversity property of the received signals, the Multiple-Input Single-Output (MISO) blind equalizer is constrained and a good initial value is sought to overcome the local convergence of the blind equalizer. Secondly, by taking the characteristics of transmitted signals, the influence of time-varying channels, and the fast optimization theory into account, we develop a fast method based on high precision dual-mode cost function and fast optimization algorithm to estimate the MISO blind equalizer. Thirdly, a recursive scheme is proposed to estimate the MISO blind equalizer. Therefore, the bottleneck that the existing algorithms are unable to estimate several MISO blind equalizers successively is broken through; meanwhile the computational complexity is reduced to a great extent. On these bases, a complete scheme of blind equalization for MIMO communication systems under time-varying channels is given, and its convergence stability, convergence speed and equalization accuracy would be evaluated by simulations, which provides a theoretical basis for its application in the time-varying MIMO communication systems.

中高速移动MIMO通信系统因高机动性和大容量的优点备受关注。盲均衡补偿信道失真的同时保证了该类系统高频谱利用率。本项目针对中高速移动条件下时变MIMO系统盲均衡方法局部收敛、缓慢收敛及高计算复杂度问题，拟运用快速递归提取方法开展时变MIMO系统盲均衡方法研究。首先，结合接收信号空时分集特性，约束MISO盲均衡器，寻求良好初值克服盲均衡器局部收敛问题。其次，通过分析传输信号特征和时变信道影响，结合快速优化理论，研究基于高精度双模式代价函数和快速优化算法的MISO盲均衡器快速估计方法。再次，突破现有算法无法依次估计多个MISO盲均衡器导致高计算复杂度的瓶颈，提出基于递归方法的MISO盲均衡器估计方案，达到MIMO系统低复杂度盲均衡的目的。在此基础上，给出完整的时变信道下MIMO通信系统盲均衡研究方案，并通过仿真评估算法收敛稳定性、收敛速度及均衡精度，为其在时变MIMO通信系统应用提供理论基础。

中高速移动MIMO通信系统因高机动性和大容量的优点备受关注。盲均衡补偿信道失真的同时保证了该类系统高频谱利用率。本项目针对中高速移动条件下时变MIMO系统盲均衡方法局部收敛、缓慢收敛及高计算复杂度问题，运用快速递归提取方法开展时变MIMO系统盲均衡方法研究。首先，利用接收信号空时分集特性，约束MISO盲均衡器，得到了良好初值，克服了盲均衡器局部收敛问题。其次，通过分析传输信号星座图特征和时变信道对接收信号的影响，结合快速优化理论，研究基于高精度双模式代价函数和快速优化算法的MISO盲均衡器快速估计方法。再次，突破现有算法无法依次估计多个MISO盲均衡器导致高计算复杂度的瓶颈，提出基于递归方法的MISO盲均衡器估计方案，达到MIMO系统低复杂度盲均衡的目的。在此基础上，给出完整的时变信道下MIMO通信系统盲均衡研究方法，并通过仿真评估算法收敛稳定性、收敛速度及均衡精度，为其在时变MIMO通信系统应用提供了理论基础。