基于遗传算法和位爬山算法的混合多用户检测

针对传统多用户检测中的遗传算法容易陷入局部最优解且收敛慢、影响系统实时性等问题，研究利用爬山算法的优势来弥补遗传算法的这一缺陷，使2种算法优势互补，并在同步CDMA系统中对其性能进行了研究．仿真结果表明，提出的这种算法能有效克服传统遗传算法的局限性问题，且计算量小，收敛速度快，在抗干扰与克服“远——近”效应方面均有明显的优势．

Mixed multiuser detection based on genetic algorithm and bit hill climbing algorithm

Since the conventional genetic algorithms for multi-user detections have the disadvantages such that they often fall into the local optimal solutions, converge too slowly and have bad real-time property, the hill climbing algorithm is employed to overcome these disadvantages. The performances of this new algorithm was studied by using the synchronous CDMA system . Simulation results show that this method can efficiently overcome the limitations of genetic algorithm in multi-user detections. Furthermore, it has simple calculation complexity and fast convergence. Also, it is effective in withstanding the disturbances and the near-far effect.