一种基于无约束最优化的多用户检测算法

A Multiuser Detection Algorithm Based on Unconstrained Optimization Approach

针对码分多址(CDMA)系统提出了一种基于无约束最优化的近最大似然多用户检测算法.最大似然检测算法可以看成是一个具有约束条件的组合优化问题,所提出的算法通过放松这个组合优化问题的约束条件,将最大似然检测算法转化成为一个无约束最优化问题,并利用无约束最优化问题的解近似得到最大似然解.数值结果显示,无论在加性高斯白噪声信道(AWGN)下,还是在随机时变多径信道下,该算法都与最大似然检测的性能十分接近,同时其复杂度在用户数比较大时仅与线性检测算法相当.

A near maximum-likelihood （ML） multiuser detection algorithm based on unconstrained optimization approach is proposed for code-division multiple access （CDMA） systems. By employing the proposed algorithm, ML detection regarded as a combinatorial optimization problem, is transformed to an unconstrained optimization problem. The detected symbols are obtained by approximating the solution of the aforementioned unconstrained optimization problem. Simulation results show that the performance of the proposed algorithm is very close to that of the ML detection algorithm over both additive white Gaussian noise （AWGN） and time-varying multipath fading channels, meanwhile its computational complexity is comparable with that of the conventional linear detection algorithms, especially when the number of users is high.