Real-time Interference Graph Building Algorithm For Cellular D2D Communication System

In recent years,the rapid development of mobile communication technology has greatly facilitated peo-ple's life.However,the traditional base station centric network structure has greatly restricted the coverage and service provision of the cellular mobile communication system and hindered the further improvement of the capacity,which cannot meet the growing needs of users.To this end,the concept of Device-to-Device(D2D)communication has been proposed and applied to such scenarios as cellular networks.Cellular D2D communication is a technology for direct communication between mobile terminals under the control of a base station which can effectively reduce the base station load and save the spectrum resources.For cellular D2D communications,resource allocation and power control algorithms are two key technologies.Many interference graph based resource allocation and power control algorithms have been proposed in the litera-ture..The so-called interference graph model is to abstract the link and link interference using the vertices and edges of the graph,respectively.Most studies assume the interference graph has been known in advance and then use the matching,coloring,or other graph theory methods to obtain the performance improvement.However,links in cellular systems are always dynamically changing,which will lead to the dynamic changes in the cell interference graph.If the interference graph cannot be updated in real time,to allocate resource based on incomplete interference graphwill lead to the performance decline.In this paper,we proposed a real-time interference graph building algorithm for the underlay deployment of cellular D2D communication system.The algorithm divides the radio resource into the control part and the data transmission part,and constructs the interference graph by using the resource pool of the control domain part,which is an iterative process.The algorithm divides the control resource pool into two parts for cellular link broadcasting and for D2D link broadcasting,respectively.The edges in interference graph are divided into three parts of CToD(cellular-to-D2D),DToD(D2D-to-D2D),and DToC(D2D--to-cellular)according to the type of transmitting end and receiving end.we proposed different building algorithms for different types of interference edges.The algorithm aims to lower the missing rate of interference edge,combines the mechanisms of minimum sum-collision and multiple feedback to build the DToD and DToD interference edges and uses the priority priority feedback-in-turn mechanism to build CToD interference edges.Then,the performance of the proposed algorithm is analyzed theoretically.According to the different types of interfer-ence edges,the theoretical values of the three types of interference edges arc analyzed.Simulation results are reported to verify the accuracy of the algorithm analysis.The main innovation of this paper is to extend the interference graph algorithm to underlay scenes,dynamic scenes,and construct a directed interference graph.Finally,we set up a series of simulation scenarios to verify the performance of our proposed algorithm.By comparing RED(Random Early Detection)algorithm,cyclic RED algorithm,minimum sum-conflict al-gorithm,no-feedback,simple feedback,and multiple feedback algorithms,and the priority feedback-in-turn algorithm,it is found that the performance of the algorithm proposed in this paper the combined algorithms of priority rotation algorithm,minimum collision,and multiple feedback are superior in real-time performance at the interference edge.Improved performance over traditional RED algorithms by more than 5 times in low-rate dynamic scenarios.By changing the values of variables,the paper extensively evaluated the impact of parameters such as the number of cellular links,number of D2D links,access speed of cellular links,access speed of D2D links,maximum D2D communication distance,and number of resource blocks on performance of the proposed algorithm.