Design Of Massive MIMO Multi-user Precoding Based On Swarm Intelligence

Massive multiple-input multiple-output(MIMO)technology can effectively eliminate inter-user interference by arranging a large number of antennas on the base station side and taking advantage of the asymptotic orthogonality between channels,and at the same time,it can bring about a huge performance improvement to the communication system,becoming a key technology of 5G has been widely used.In a massive multi-user MIMO(MU-MIMO)system,incorrect user grouping will cause severe inter-user interference,so multi-user selection is critical to the performance of the massive MIMO system.At the same time,in massive MU-MIMO systems with massive users,the traditional multi-user selection algorithm based on exhaustive search has extremely high computational complexity,and the previously proposed low-complexity greedy multi-user selection algorithm has large performance loss.Therefore,this paper studies the multi-user selection problem to solve the above problems,and the specific work is as follows:(1)The population-based incremental learning(PBIL)algorithm was utilized to propose a PBIL-based MU-MIMO multi-user selection algorithm in this paper.Firstly,the initial probability vector was used to generate the initial population.Secondly,the best individual and dominant individuals were extracted from all individuals through competitive learning,and their information was used to update the probability vector.Then,an orthogonal arrangement was designed to mutate the best individual in order to increase the diversity of solutions,and the updated probability vector and orthogonal arrangement were utilized to generate new individuals in the next-generation population,so as to realize the evolution of the population.Finally,the information entropy of the system was introduced as the termination condition of the evolution,and the multi-user selection scheme was output after the evolution was completed.The proposed algorithm is not only efficient,but also can avoid the local trap problem of the classic PBIL algorithm.Simulation results show that in the case of massive connections,the proposed algorithm can obtain good system performance while maintaining low complexity.(2)The PBIL based on decimal coding(D-PBIL)algorithm was introduced to propose a user-centric D-PBIL-based cell-free massive MU-MIMO access point(AP)selection algorithm in this paper.Firstly,the initial probability matrix was used to generate the initial population.Secondly,the best individual was extracted from all individuals through competitive learning,and their information was used to update the probability matrix.Then,a mutation method combining directed mutation and random mutation was designed to speed up the convergence of the algorithm and increase the diversity of individuals.Finally,after the population evolution was completed,the information entropy of the system was used as the termination condition of the iteration,and the AP selection scheme was output after the iteration.The cell-free massive MIMO system uses a large number of low-cost service antennas called APs to provide services to users distributed in the communication coverage area,which can improve the spectrum efficiency of 95% of the users in the system,but this performance advantage comes at the cost of huge backhaul overhead and power consumption.Simulation results show that the designed algorithm can greatly reduce the power consumption of the cell-free massive MU-MIMO system while achieving good performance,and to a certain extent increase the fairness of users being served.