Deep Reinforcement Learning Based Mechanism For Resource Allocation With High Energy Efficiency In 5G Network

With the rapid development of Internet,a variety of terminal devices are growing rapidly.In order to cope with the challenges brought by the rapid growth of the number of terminal connections and mobile data traffic,and to provide users with better use experience,lower transmission delay and higher transmission rate services,5G network came into being.The application of 5G network brings new development direction and prospect to more business scenarios,such as autopilot,mass machine communication,edge computing,industrial Internet of things(IIoT)and other typical business application scenarios,which also meet greater development opportunities due to the development of 5g technology.In order to save energy and reduce cost,energy efficiency(EE)in 5G network is also the focus of academia and industry.Therefore,with the rapid development of 5G,in the face of a large number of end users and diverse requests,how to meet the business service quality and allocate resources reasonably and efficiently has become one of the urgent problems to be solved.In the existing work of 5G network resource allocation,there are some problems such as slow allocation speed and low accuracy.Deep reinforcement learning(DRL)provides a better solution to complex problems.At present,it has achieved remarkable results in many fields.However,in the application of 5G network resource allocation,the consideration of energy efficiency is not deep enough,and the quality-of-service requirements of users are not detailed.Edge computing and IIoT are two typical application scenarios in 5G network.Therefore,this paper studies the resource allocation methods for these two typical scenarios,and proposes an energy-efficient resource allocation method based on DRL algorithm,which improves the network performance and reduces the network operation cost.Aiming at the super HD video transmitting service scenario of edge computing,this paper proposes a resource allocation method for 5G high bandwidth service based on deep Q network(DQN).Firstly,according to the characteristics of edge computing,a resource allocation model for quality of service(QoS)guarantee is designed by taking the connection relationship between base station and users and the transmission power allocated by base station to users as decision variables,minimizing the overall energy efficiency as the goal,and taking the needs of mobile users as constraints.On this basis,a resource allocation framework of edge computing based on DQN is proposed.Convex optimization algorithm is used to obtain the minimum transmission energy under a certain connection relationship,and then DQN is used for iteration.On the basis of convex optimization results,the optimal connection relationship and optimal power allocation value are found to calculate the optimal energy efficiency value.Finally,the simulation results show that the proposed method has faster training speed and convergence speed,ensures the quality-of-service requirements of mobile users,and realizes intelligent and energy-saving resource allocation.For IIoT environment,this paper proposes a resource allocation method for 5G wide connection low delay services based on asynchronous advanced actor critical(A3C).Firstly,according to the characteristics of IIoT environment,the connection relationship between the base station and the user and the transmission power allocated by the base station to the user are taken as the decision variables,and the energy efficiency maximization while meeting the needs of each user is taken as the optimization objective.On this basis,an energy-efficient resource allocation method based on A3C is proposed.The hierarchical agglomerative clustering(HAC)algorithm is used to determine the connection relationship between the base station and users,and the A3C algorithm is used to allocate transmission power to users,so as to maximize the overall energy efficiency and ensure the demand of each user.Finally,the performance of different reward functions,different resource allocation methods and different clustering algorithms is verified through simulation experiments.It is proved that the proposed method has faster convergence speed and the result is closer to the optimal solution.The goal of maximizing energy efficiency and meeting the QoS requirements of each user is achieved.To sum up,this paper proposes and implements a 5G access network energy-efficient resource allocation method based on deep reinforcement learning.According to the characteristics of different 5G application scenarios,the corresponding models and frameworks are proposed,which provides a new idea and method for solving the problem of energy-efficient resource allocation in different scenarios.