Research On Dynamic D2D Relay Selection Algorithm Based On Social Relationship

In recent years,with exploiting widely of emerging technologies such as video streaming,mobile computing,and virtual reality games,critical requirements for mobile real-time bandwidth are proposed.Since current 4G cellular communication technologies cannot carry the dense connections of massive cellular users,it is necessary to find new communication technologies to promote the development of cellular communication networks.Device-to-Device(D2D)communication is the new type of communication technology,and it is a strategic technology in 5G mobile communication.D2 D communication in a cellular network refers to direct communication between two mobile users without traversing the base station(BS)or core network,which actually is a short-distance communication technology.In cities,D2 D transmission signals may be greatly attenuated due to the occlusion of buildings,deep channel fading,and co-channel interference from cellular devices.Therefore,in many cases,if two users establish D2 D direct communication,the transmission performance between them would be poor,and communication may even be interrupted.D2 D relay communication refers to making use of D2 D relay nodes to forward data sent from the source to the destination.By introducing a relay node,it is possible to amplify the attenuated signal,increase the transmission rate of the system,expand the communication transmission range,optimize the utilization efficiency of spectrum resources.The advantages of D2 D relay communication make it very important to choose a relay node.This paper mainly explores the selection algorithm of D2 D relay selection scheme in two main mobile application scenarios.The specific research content is:Aiming at the problem of D2 D relay selection among users in fast-moving application scenarios,in order to obtain efficient and stable relay-assisted communication,a dynamic D2 D relay selection scheme based on relative speed and social relationship is proposed.First,the solution initially divides candidate relay nodes based on on the characteristics of the physical domain.Then,social relationships and social interests are used to define the social strength which represents the willingness to relay auxiliary communication.At the same time,it analyzes the factors that affect communication stability,a communication stability formula is proposed,and the obtained candidate relay nodes' communication stability is arranged in reverse order.Finally,the optimal stopping algorithm is used to select the optimal relay-assisted communication from the candidate relay nodes to obtain a stable relay node that balances the detection cost and system performance.Through simulation analysis,it is verified that the proposed algorithm can achieve greater throughput,stability,success rate and smaller communication delay.Aiming at the problem of D2 D relay selection of a dense user scenario in a slow-moving environment,in order to obtain more intelligent relay-assisted communication,and a relay selection program based on Deep Q Network(DQN)is proposed.First,this solution screens candidate relays for D2 D users based on the social intensity,the communication distance between devices,the remaining power of the devices,and the remaining cache space.Then,state variables are constructed of the factors affecting the relay selected which can use as training data to train the network.Secondly,define the reward function consisting of cellular user throughput,D2 D user throughput,social intensity among D2 D users,remaining power of the device,and cache.Finally,with the goal of maximizing the expected value of the cumulative discount reward,the network training action is taken as the selected relay user and the matching channel resource.The simulation results show that the proposed algorithm can improve the throughput of the system and guarantee the success rate of communication.