中文摘要：

基于动态子信道分配的MIMO-OFDM波束成形系统可以获得比传统系统更好的误码率性能。但由于处理或者反馈时延的影响,会使得发射机利用不准确的信道状态信息发射数据,从而在数据流之间引入干扰,导致系统性能下降。通过分析时延对系统的影响,该文提出了一种改进的检测算法：通过构建等效MIMO系统模型,采用最小均方误差MIMO检测算法抑制数据流之间的干扰,并利用等效MIMO信道矩阵的特征值分解白化噪声获得等效高斯信道模型,计算编码比特的对数似然比。仿真表明,该算法相对于现有算法有2.5～4dB的性能增益。

英文摘要：

Dynamic subchannel allocation based MIMO-OFDM beamforming system outperforms conventional MIMO-OFDM beamforming system with considerable gain. Unfortunately, the imperfect channel state information, which is due to processing or feedback time delay, will be used by the transmitter to transmit data symbols. The resulted inter-streams interference will degrade the system performance significantly. In this paper, an improved detection scheme is proposed to combat effectively with this interference according to the analysis of the effect due to delay. By constructing equivalent MIMO system model, the inter-streams interference is suppressed though MIMO detection algorithms. Furthermore, eigenvalue decomposition of equivalent MIMO channel is applied to whiten the noise so that equivalent Gaussian channel model can be obtained to compute the log-likelihood ratio of coded bits. Simulation results show the proposed detection scheme outperforms conventional scheme with a gain of 2.5-4 dB.