几种 MIMO 最大似然检测算法性能与复杂度比较及改进，介绍了半定松弛、分枝定界和堆栈三种低复杂度最大似然检测算法，并对其性能和复杂度进行了仿真分析，提出了改进的分枝定界和堆栈算法，仿真结果证明分枝定界和堆栈算法性能要优于半定松弛算法，分枝定界算法的复杂度低于堆栈算法且半定松弛算法以多项式复杂度取得了逼近最大似然的性能，同时改进算法加快了算法收敛速度，降低了计算复杂度和对存储空间的要求。

-Several maximum likelihood MIMO detection algorithm performance and complexity comparison and improvement, introduced the semidefinite relaxation, branch and bound, and the stack of three low-complexity maximum likelihood detection algorithm and its performance and complexity of the simulation analysis , proposed an improved branch and bound, and the stack algorithm, simulation results show that the stack algorithm branch and bound and given a better performance than a half relaxation algorithm, branch and bound algorithm complexity is lower than the stack algorithm and the algorithm for semidefinite relaxation polynomial complexity approximation of the maximum likelihood achieved performance, while improving the algorithm to speed up the convergence rate and reduces the computational complexity and storage space requirements.