基于预编码矩阵的迭代均衡算法

Iterative equalization algorithm based on precoding matrix

针对长期演进(long time evolution,LTE)下行多输入多输出正交频分多址链路(multiple-input multiple-output orthogonal frequency division multiplexing,MIMO-OFDM)异步通信系统中的天线间干扰和多径干扰的问题,提出一种低复杂度的基于预编码矩阵的迭代均衡算法。在发射端,该算法通过预编码矩阵将信号扩展到所有子载波上,从而降低部分子载波深衰落对扩展前原始信号的影响。在接收端,利用最小均方差误差排序QR分解(minimum mean square error sorted QR decomposition,MMSE-SQRD)软输入软输出干扰消除均衡算法,一方面避免传统基于最小均方误差(minimum mean square error,MMSE)并行软干扰消除均衡算法中复杂的矩阵求逆运算,进而降低了算法复杂度,另一方面利用信道排列优先检测信噪比最大的传输符号提高检测准确性。同时通过预编码对重构信号中误差进行扩展,进而缓解在迭代干扰消除过程中的误差传播。仿真结果证明,在2发2收场景下,误码率在10-3时,算法经过5次迭代后系统性能相比于现有的迭代均衡算法改善约4dB。

Aiming to co-antenna interference and multipath interference in the LTE downlink multiple input multiple output orthogonal frequency division multiplexing （MIMO-OFDM）asynchronous communication sys-tem,a low complexity iterative equalization algorithm based on precoding matrix is proposed.At the transmit-ter,the algorithm utilizes precoding matrix to spread the symbols over all the subcarriers,which relieves the in-fluence of deep fading in part of the subcarriers.At the receiver,the algorithm adopts soft input soft output in-terference elimination algorithm based on the minimum mean square error sorted QR decomposition （MMSE-SQRD）to avoid the complexity of solving inverse matrix,which is always involved in the traditional parallel soft interference elimination algorithm based on minimum mean square error equalization.Utilizing a sorted QR de-composition of the channel matrix,on the other hand,the algorithm has the priority for detection transmission symbol with maximum noise-signal ratio to improve detection accuracy.Furthermore,the error of reconstructed signals is spread by the precoding matrix,which decreases the error propagation in iteration processing.Simulation re-sults show that the performance of the proposed algorithm is improved compared with the existing iterative interference cancellation algorithm,that is,when the system is equipped with 2 transmitters and 2 receivers and the bit error rate is 10 -3 ,the performance of the system is improved about 4 dB.