中文摘要：

针对非理想信道状态信息下两跳中继多输入多输出（MIMO）系统，按照空间相关性进行了非理想MIMO信道建模，在此基础上提出了一种新的多节点联合优化设计该系统收发信机的算法。该算法基于估计信号的最小均方误差（MMSE）准则，把源节点预编码、中继节点预编码以及目的节点均衡算法结合起来进行三节点联合优化，提高了通信系统的可靠性。收敛性和复杂度的仿真分析表明，该三节点联合优化算法与传统的中继和目的节点两节点迭代联合优化算法相比较，误码率有1dB以上的增益，复杂度大致相当；如果该算法退化为两节点的话，虽然在性能上与文献给出的方法大致相当，但是复杂度可以降低30％左右。该算法复杂度的降低将会以占用存储资源为代价，因为在计算过程中很多矩阵会被重用，中间变量需要存储下来，但从综合衡量性能与复杂度的角度看，该方法具有一定程度的优势。

英文摘要：

A novel transceiver joint optimization algorithm was proposed for dual hop multiple input multiple output （ MI- MO） relay communication systems with imperfect channel state information （CSI）. Firstly, an imperfect MIMO channel model was built according to spatial correlation property, and then, based on the minimum mean square er- ror （MMSE） criterion, this algorithm was optimized by jointly combining source node precoding, relay node pre- coding and destination node equalization together, which ensured the reliability of communication systems. The convergence and complexity were analyzed by simulation and the results were given. Compared with the traditional two nodes iterative optimization algorithm at relay and destination nodes, if the proposed algorithm is sufficiently convergent, it＇ s BER performance will be improved 1 dB or more, while the complexity is almost the same, and when the algorithm is changed from three nodes to two nodes joint optimization, though the performance is almost unchanged, while the complexity will be reduced about 30%. Meanwhile, with this proposed algorithm, there are many repeated intermediate variables such as matrix multiplying results need to be stored in memory before band, the complexity reduction is at the cost of the memory utilization. But when the performance and the complexity are considered, the proposed solution is better than the two nodes joint iterative optimization solutions existing at pres- ent.