Research On Resource Allocation Of D2D Communication In Cellular Network

With the rapid spread of smart terminals and their applications,the problem of a surge in data traffic has attracted great attention.In order to improve spectral efficiency,device-to-device(D2D)communication technology is proposed.Initially,D2 D systems are mainly deployed in licensed frequency bands to communicate by reusing the channel resources of the cellular system.With the increase of mobile users,the licensed frequency band has become more and more crowded,and D2 D has therefore begun to be deployed in the unlicensed frequency band(D2D-U).However,the introduction of D2 D technology in the unlicensed frequency band will affect the communication quality of WiFi users,so it is necessary to consider the issue of fair coexistence with the WiFi system.Therefore,this thesis mainly studies the resource allocation of D2 D communication deployment in licensed frequency bands and unlicensed frequency bands.1.Considering that D2 D communication is deployed in unlicensed frequency band,and the duty cycle is adopted to coexist fairly with the WiFi system,this thesis proposes a D2D-U resource allocation algorithm based on Q-learning.In this algorithm,the small base station acts as the agent.The actions of the agent are defined as different combinations of the transmit power and duty cycle of the D2 D user,and the states of the agent are defined as different combinations of the signal-to-noise ratio of the LTE-U user,the throughput and the fairness of the coexistence system.By constantly interacting with the environment,the small base station can train a stable Q table.And it can select the optimal transmit power and duty cycle allocation for the D2 D user through the Q table.The simulation results show that the algorithm has better throughput and fairness of the coexistence system than the traditional algorithm.2.Considering the joint deployment of D2 D communications in licensed and unlicensed frequency bands,this thesis proposes a resource allocation algorithm with joint power control,uplink and downlink channel allocation and mode selection.In the licensed frequency band,D2 D users can reuse the uplink and downlink channels of cellular users.And in the unlicensed frequency band,D2 D users use duty cycle mechanism to coexist with WiFi system.Firstly,the problem model of maximizing the throughput of cellular users and D2 D users is established on the premise of ensuring the communication quality of each user,and then the optimization problem is decoupled into three sub-problems of power control,uplink and downlink channel allocation and mode selection.Finally,the geometric programming method is used to solve the power control problem,and the Hungarian algorithm is used to solve the uplink and downlink channel allocation problem.And in the mode selection,D2 D users who cause severe interference to cellular users are connected to the unlicensed frequency band.The simulation results show that the algorithm has better system throughput.