Research On Interference Management Technique In Ultra-dense Networks

To cope with the explosive growth of data traffic demand for 2020,the fifth generation of mobile communication technology arises at the historic moment.As one of the key technologies of 5G,Ultra-Dense Networks(UDN)technology can effectively improve the power efficiency,spectrum efficiency and system capacity of the whole network.However,even the total increased cells in UDN can enhance the entire network system capacity,UDN will face many new technical challenges,such as a variety of interference coexistence,mobility switching problems with the increase of community deployment of density.Interference Alignment(IA)technology was honored with its unique advantages in numerous interference management strategy.However,using traditional IA in UDN directly to eliminate interference of all base stations will face a great deal of signaling overhead and the high iteration complexity.How to effectively solve the bottleneck problem is still the main research focus in the academic circles.Therefore,the research significance of this paper is based on the spatial random geometric distribution poisson point process Voronoi network model to construct UDN networking architecture.Developing more effective clustering strategies combined with interference alignment technology to enhance the overall capacity of the UDN system.The innovation point of this article is based on the idea of clustering interference alignment.It puts forward two clustering algorithms based on graph partition.The whole network is divided into several clusters,to achieve the goal of eliminating the strong interference within the clusters.First strategy based on Distance clustering scheme focuses on the implementation of the low complexity.The selection of cluster heads is based on the near neighborhood average distance and the choice of nodes into the cluster based on the index of Distance.The number of communities in each cluster is constant.The formation structure of the last cluster may not be reasonable to limit the overall rate of the system.The second one is based on Neighbor Interference Ratio(NIR)node selection which focuses on the more reasonable cluster structure and realizes better system performance.The development of the index NIR takes into account the factors of distance and interference.The selection of cluster heads is based on the mean of adjacent interference,and the selection of cluster nodes is based on NIR index,and the structure of the last cluster is optimized by certain threshold value.Therefore,this scheme allows the number of communities in the cluster to change dynamically within a certain range.A more reasonable clustering result and a better system rate performance improvement can be obtained,but the complexity is slightly higher than that of the first clustering scheme.The simulation results show that the proposed strategy can effectively suppress.