| 5G is the latest generation of cellular mobile communication technology,which has the advantages of high data,high speed,low delay and low consumption.It will become the basic general technology of the new generation following steam engine,electric power technology and Internet technology,leading the fourth industrial revolution.Massive MIMO,as a key technology of 5G,has excellent beam forming capability,effectively improving coverage capability,reducing interference within the system,providing higher spectrum efficiency and effectively improving user experience.There are thousands of kinds of Massive MIMO beam combinations.If the Massive MIMO beam is not applied properly,it will bring new problems such as beam interference,complex parameter configuration,scene matching and so on.Starting from the basic theory of Massive MIMO beam,based on the coverage scenarios and UE distribution,this paper utilizes the expert experience strategy,iterative optimization,reinforcement learning and other methods to produce the optimal beam forming by using the flexible antenna weight setting of Massive MIMO.Massive MIMO beam adaptive adjustment is realized by using MR+MDT+KPI network management data of 5G cell standard to extract user and business characteristics.Massive MIMO single cell coverage and capacity can be dynamically matched through dynamic adjustment of antenna weight assignment and beam splitting scene and time division configuration.On the basis of single cell optimization,according to the cell topology structure,user and business distribution,through iterative intelligent optimization algorithm,the whole network Massive MIMO antenna weight optimization is implemented to achieve the overall optimal performance of wireless network.The Multi-cell weight optimization is a combinatorial optimization problem,the number of weight combinations grows exponentially with the increase of cells,and the optimization process requires a great deal of calculation.In Massive MIMO weight optimization,artificial bee colony algorithm and genetic algorithm are combined.Firstly,artificial bee colony algorithm is used to generate a relatively well-distributed initial population,and then genetic algorithm is used to solve the problem.The two algorithms complement each other.Due to the good distribution of the initial population,the pressure of genetic algorithm in crossover and mutation will be relatively low.The local searching ability of artificial bee colony algorithm is relatively strong,which can maintain the diversity of population,reduce the possibility of the algorithm falling into the local optimal solution,make the algorithm easier to search the global optimal solution of the problem,and accelerate the convergence of the algorithm.After the optimization of antenna weight in the current network 5G community,the evaluation results are as follows: the average value of CSI-RS RSRP is basically the same,and the proportion of medium good points is increased by 1%;the average value of CSI-RS SINR is increased by 1.34 d B,and the proportion of medium good points is increased by 1.69%.The average value of SSB RSRP remained the same,and the proportion of medium and good points increased by 2.12%;the average value of SSB SINR increased by 2.16 d B and the proportion of medium and good points increased by3.4%;the average downlink download rate increased by 129 Mbps,with an increase rate of18.35%.Through 5G antenna adaptive beam forming,the coverage of 5G cell can be matched with user and business distribution,so as to effectively improve the user perception in hot spots and improve the overall throughput of 5G cell.Weight optimization replaces tower site adjustment to realize remote RF optimization,which improves optimization efficiency and saves the cost of planning,construction,maintenance and optimization. |
PDF Full Text Request