格基约减辅助的低复杂度列表检测

A Lattice Reduction-Aided Low-Complexity List Detection

格基约减作为一种矩阵近似正交化方法,能够显著改善传统MIMO检测算法的性能。推导了格域星座点各分量之间的递推约束关系,并提出了格基约减辅助的列表SML-SIC检测算法,相比格基约减辅助的列表SIC检测,获得了更好的性能。在所提算法中,列表检测的子检测器为SML-SIC检测,它是将2维的列表检测器(SML)与连续干扰抵消(SIC)检测方法相结合,基于串行干扰抵消思想,用SML检测器每次对两个符号进行检测,获得了比SIC检测更高的分集增益。另外,提出局部格基约减辅助的列表检测,选择列表长度等于调制阶数,并对较低维的信道矩阵进行格基约减,复杂度有所下降,但获得和格基约减辅助的列表检测相同的性能。

Lattice reduction, as an approximate matrix orthogonalization method, could markedly improve the performance of traditional MIMO detection. The recursive constraint relations among the components of constellation point in lattice domain are derived, and a lattice reduction aided list SML-SIC detection algo- rithm also proposed, this algorithm enjoys better performance as compared with lattice reduction aided list SIC detection. In the proposed algorithm, as the sub-detector of list detection, with combination of 2d list detector （SML） with SIC detection method, and based on the idea of successive interference cancellation, SML-SIC algorithm could detect two symbols each time by using SML detector, and acquire higher diversi- ty gain than SIC detection. In addition, a partly lattice reduction aided list detection is proposed, the length of list is selected to be equal to modulation order, and a lower dimensional channel matrix is pro- cessed in lattice reduction, thus to decrease the complexity while maintaining the identical performance with lattice reduction aided detection.