中文摘要：

针对虚拟多天线正交频分复用（VMA-OFDM）系统目的节点中数据辅助的多频偏和信道参数估计问题进行了研究,设计了一种次优最大似然估计（MLE）算法.该算法在矩阵的求逆中采用了级数展开,从而使得算法的复杂度与基于最大似然准则估计算法相比显著减小,在算法的性能和复杂度之间进行折衷,解决了最大似然估计算法过于复杂而难以实现的问题.在算法的性能方面,理论分析和仿真结果表明,基于该算法的频偏和信道参数估计的均方误差逼近Cramer-Rao下界（CRB）;在算法的效率方面,该算法允许各个中继节点同时发送训练序列,从而降低传输训练序列所占用的时间.

英文摘要：

The joint data-aided estimation of multiple frequency offsets and channel coefficients in virtual multi-antennas orthogonal frequency-division multiple（VMA-OFDM） systems were studied, and a suboptimal maximum likelihood estimators（MLE） was introduced. The series was expanded in the matrix inversion. Compared with the existing estimators based ML, the computational cost of the proposed estimators is reduced significantly. The algorithm exhibits an attractive tradeoff between performance and complexity, and properly settles the problem that the MLE is impractical in this context. As far as the performance is concerned, the proposed suboptimal estimators are shown to be asymptotically efficient, i. e. , the frequency offsets and channel coefficients estimation mean square error（MSE） achieves the Cramer-Rao bound（CRB）. Simulation results sustain our claims. As far as estimation efficiency is concerned, the algorithm allows the relay nodes to transmit the training sequence simultaneously, which will greatly reduce the training sequence transmission time.