CoMP下行传输系统的联合最优参考序列设计
Joint Optimal Training Sequence Design for CoMP Downlink System
-
摘要: 针对多小区协作传输系统, 首先给出了频偏估计和信道估计的Cramer-Rao Bound (CRB) , 作为理论上最优的训练序列的设计准则;然而通过分析发现, 基于CRB的最优参考序列设计准则中不仅包含了信道的因素, 而且还包含了协作基站频偏值的影响.由于基于CRB的最优参考序列设计过于复杂, 因此转向次优的基于最小均方误差 (MSE) 的参考序列设计准则, 并通过分析得出该准则下的最优参考序列是高斯序列.最后通过仿真, 验证了高斯序列在最大似然估计下的性能.仿真结果显示:高斯序列检测性能逼近理论最优值CRB;通过与其他序列的检测性能进行比较, 验证了其性能增益.Abstract: In this paper, the Cramer-Rao Bound (CRB) to frequency offsets and channel impulse response estimation of Coordinate on Multiple Point downlink transmission system (CoMP) was derived for optimal training sequence design at first. However, the optimal sequence design criterion based on CRB was too complicated as the factor of channel impulse response and frequency offsets can't be eliminated in this criterion. So a sub-optimal design criterion based on minimum Mean Square Error (MSE) was proposed, then it was found out that the gauss sequence was the optimal sequence under this criterion.The numerical simulation verifies the theory analysis under the maximum likelihood (ML) estimation.The results show that the estimation performance approach CRB quiet well and gauss sequence outperform other sequences.