针对存在空域相关性的多输入多输出(Multiple—Input Multiple—Output,MIMO)信道,首先设计了适用于任何给定长度的最优训练序列,该序列能够充分利用已知的信道相关信息,且具有估计均方误差最小化的特性.在此基础上,分析了相关信道下采用训练估计的数据传输机制所能达到的容量下界,得出了满足估计准确性和数据速率最优折中的最佳训练序列长度.最后,通过仿真验证了本文给出的训练序列的性能和最佳长度取值的有效性。
The optimal design of training sequences for Multiple-lnput Multiple-Output (MIMO) channel estimation with spatial correlation is considered. Based on the criterion of Minimum Mean Square Error (MMSE), the optimal training sequences that have controllable length are designed to exploit full information of the spatial correlation. Furthermore, a lower bound on the capacity of a waining-based MIMO system is derived, from which the optimal number of training symbols is given, It is investigated that the proposed training sequences have good performance and the capacity lower bound is valid.