Research On The Resource Allocation Strategy Based On Clustering In Ultra-Dense Network

Driven by the exponential increasing of traffic,the cellular networks are developing towards the direction of miniaturization and eventually to Ultra-Dense Network(UDN)to meet the high network capacity and spectrum efficiency requirements in hotspots.With the densification of micro-base stations,the interference within the system becomes more complicated.How to deal with the high computational complexity and signaling overhead in the frequency and power allocation algorithms in the large-scale network topology is the most urgent problem to be solved in the practical engineering application of ultradense network.The purpose of this paper is to reduce the scale of frequency and power resource allocation problem by studying the clustering algorithm of micro-base stations in ultradense network,so as to avoid the cost of solving large-scale problems and make it more engineering achievable.At the same time,efficient frequency and power allocation strategies are designed to improve system throughput and user service quality,and to optimize the convergence performance of the algorithm.Firstly,this paper describes and summarizes the research status of clustering and joint distribution of frequency and power in ultra-dense networks.The system architecture and system characteristics brought by network intensification are analyzed.Then this paper analyzes the changes of system architecture and system characteristics brought by network densification,and the system simulation model is built,including site model and channel model.Besides,it analyzes the influence of microbase station density on coverage area and utilization rate of microbase station,laying a foundation for the following research.Then,in order to reduce the scale of the optimization problem,this paper studies the clustering algorithm of micro-base station.The micro-base stations in the system are divided into cluster base stations and split base stations for isolating the inter-cluster interference of the cluster base station and the split base station first allocates frequency resources different from the cluster base station,and forms an interference buffer zone between the clusters of the cluster base stations.Then the cluster base stations reuse the remaining frequency resources and performs frequency resource allocation among clusters and within the clusters respectively according to the principle of minimum cochannel interference,wherein the resource allocation process in each cluster of clusters can be performed in parallel,and the simulation results show that the proposed The algorithm approaches the simulated annealing algorithm in terms of system throughput,which is not far from optimal performance.Finally,for the problem that there are too many users in the system that cannot meet the quality of service after frequency allocation,the problem model of frequency and power allocation and the influence of different power allocation on system performance are firstl analyzed.Then,an improved power distribution method based on genetic algorithm is designed.The algorithm completes the selection of the power allocation strategy for the co-channel interference users in the system and can effectively improve the proportion of users in the system that meet their demand at a very low iteration cost,and the convergence performance of the algorithm is better than that of the simulated annealing algorithm based on global information.