研究无线信道优化问题, 由于非确定性 M MIO系统的信道矩阵行不满秩特性, 导致 MM SE- GDFE左预处理后引入了强干扰, 使得传统 K- Best算法产生了误码平层效应。为了消除误码平层效应,提出部分最大似然和 K - Best结合算法, 并在初始检测的强干扰区域利用部分最大似然算法获得局部最优解, 降低强干扰对系统的影响消除误码平层效应, 进而在局部最优解的基础上利用 K- Best算法完成剩余符号的检测, 保证 M MIO 检测器的软输出特性。仿真结果表明算法消除了传统 K- Best算法中的误码平层效应, 保证了检测的可靠性-Research on wireless channel optimization problem, since non-deterministic channel matrix rows dissatisfaction rank M MIO system characteristics, resulting in MM SE-GDFE left after pretreatment introduced strong interference, the traditional K-Best algorithm to generate the error floor effect. To eliminate the error floor effect, made part of the maximum likelihood and K- Best combination of algorithms and strong interference detection using part of the initial area of ​ ​ the maximum likelihood algorithm to obtain local optima and reduce the impact of strong interference elimination system error leveling effect, and then use K-Best detection algorithm to complete the remaining symbols on the basis of the local optimal solution to ensure that M MIO soft detector output characteristics. Simulation results show that the algorithm eliminates the error floor effect of traditional K-Best algorithm to ensure the reliability of detection