去蜂窝大规模MIMO系统中基于树种二进制差分进化的接入点选择算法

Tree Seed Binary Differential Evolution Algorithm Based Access Point Selection Strategies in Cell-Free Massive MIMO Systems

在去蜂窝大规模多输入多输出(Cell-Free Massive MIMO)系统中,大量接入点(Access Point,AP)同时为多个用户服务的连接方式会导致较大的功率损耗和回程链路开销。为了给用户选出最佳服务AP集合,本文提出了一种基于树种二进制差分进化的AP选择算法。首先,提出基于二进制差分进化的AP选择算法,通过多个个体的进化实现高维数据搜索的全局优化。其次,针对传统二进制差分进化算法容易陷入局部最优的问题,进一步给出基于树种优化的双机制搜索策略,利用搜索趋势(search tendency,ST)实现全局搜索和局部搜索的最佳平衡。最后,通过定义交叉率(crossover rate,CR)自适应递减准则,加快算法收敛速度。仿真结果表明,与现有算法相比,所提出的算法可显著提高系统和速率。

In cell-free massive MIMO systems,The connection approach where a large number of access points(APs)serve numerous users suffers from high power consumption and backhaul link overhead. This paper proposes the tree seed binary differential evolution algorithm based AP selection strategies to select the optimal serving AP sets. Firstly,an AP selection algorithm based on binary differential evolution(BDE)is exploited via the evolution of multiple individuals to resolve the global optimization of high-dimensional data. Secondly,a dual-mechanism search strategy based on the tree seed optimization is proposed to avoid trapping in the local optimal solution,in which search tendency(ST)is used to achieve the best balance between the global and local search. Finally,an adaptive decreasing criterion of crossover rate(CR)is defined to accelerate the convergence. Compared with the existing algorithms,simulation results demonstrate that the proposed algorithm can significantly improve the sum rate of system.