针对 Turbo- MIMO 系统中基于最大后验概率( MAP) 的软解调方案复杂度过高的问题，基于准最大似然相移键控( QML- PSK) 检测以及比特级多流编码分层调制空时( LST) 传输的思想，提出了一种适用于任意正交幅度调制( QAM) 星座图且可逼近 MIMO信道容量的低复杂度软解调算法及其接收机模型．该算法将求解外信息时的 MAP 问题转换成最大似然( ML) 解调问题，然后使用高效的 QML- PSK 检测算法求解．然而 QML- PSK 检测算法仅适用于 PSK 信号的软解调，因此文中提出了一种 QAM- PSK 软解调算法，采用比特级多流编码 LST 传输技术将高阶 QAM 信号转化为 PSK 信号．理论分析和仿真结果表明，与现有最优的基于球形译码的软解调算法相比，QAM- PSK 软解调算法能以不到其 1/2 的复杂度获得几乎相同的误码率性能，并且其低复杂度优势在问题维数增大时更为明显．

 In order to overcome the overhigh complexity of MAP- based soft demodulation strategy for Turbo- MIMO system,this paper introduces the concepts of quasi- maximum- likelihood phase- shift- keying (QML- PSK) detection and bit- level multi- stream coded layered space- time (LST) transmitter,and proposes a low- complexity soft demo-dulation algorithm and the corresponding receiver,which are suitable for all kinds of quadrature amplitude modula-tion (QAM) constellations and can approach the MIMO channel capacity.In this algorithm,the MAP problem of solving the extrinsic information of bits is converted into maximum likelihood demodulation problems that are then efficiently solved by using the QML- PSK detector.Moreover,as the QML- PSK detector is suitable only for the soft demodulation of PSK signals,a QAM- PSK soft demodulation algorithm is further proposed,which introduces the bit- level multi- stream coded LST transmitter to convert the QAM signal into PSK signals.Theoretical analysis and simulation results show that the proposed QAM- PSK soft demodulation algorithm achieves almost the same BER per-formance with a complexity less than half of that of the existing optimal sphere decoding- based soft demodulation algorithms,and that the advantage of low complexity of the algorithm is more obvious in larger problem dimensions.