Research On Resource Allocation Technology Based On NOMA In Mobile Edge Computing Scenario

With the rapid development of the global digital economy,the demand for higher communication quality is growing in various industries.Both the quantity of diversified multimedia services and the number of smart devices are increasing exponentially year by year,which brings huge data traffic and network access to the wireless networks and also encourages the innovation of communication technology.However,traditional orthogonal multiple access(OMA)technology can no longer meet the demand for massive connections in the fifth generation(5G)mobile networks.Among the candidate access technologies of 5G,non-orthogonal multiple access(NOMA)technology is considered to be a new generation of mobile communication access technology with great potential.Compared with OMA technology,the advantage of NOMA technology is that multiple users can reuse the same time-frequency resource block,which can not only increase the number of user access but also can improve the utilization of spectrum resource.Therefore,NOMA technology is extremely suitable to meet the challenges of massive equipment connections and higher quality communication in the future.With the enrichment of mobile applications,more and more mobile applications have higher requirements for the delay and computing power,such as virtual reality(VR),augmented reality(AR),automatic vehicles to vehicles,and so on.However,mobile devices are still limited by the development of the hardware and battery capacity,which will affect the performance in processing such computational intensively applications.Mobile edge computing(MEC)technology can effectively solve the above problems.MEC means to set the cloud server at the base station(BS)in the cellular cell.As the characteristic of MEC,mobile devices can offload their computing tasks to the MEC server to save delay and the energy consumption of users rather than sending the computing tasks to the remote cloud for processing.Specifically,current researches always pay more attention to the problem of transmission delay and energy consumption in the MEC downlink scenario but ignore the optimization of wireless transmission mode.Accordingly,this paper studies the NOMA-based user pairing and power allocation scheme in the MEC scenario.Based on the above analysis,this paper studies the user pairing and power allocation scheme based on NOMA in the scenario of MEC.In the NOMA system,user pairing and power allocation are two key factors that would affect the performance of the system.Due to these two factors are coupled in joint optimization problems,the problem is divided into two sub-questions to solve.In the first stage,this paper proposes a downlink NOMA system user pairing scheme based on a deep reinforcement learning(DRL)algorithm in the scenario of MEC,to maximize the sum rate of all users in a cellular cell,the user pairing problem is firs t transformed into a DRL model,then,the user pairing matrix is built according to the channel condition information of all users.Finally,the optimal user pairing scheme is found through the deep Q network(Deep Q-Network,DQN)algorithm.Simulation results show that the proposed scheme can effectively improve the sum rate of all users.For the problem of power allocation in the NOMA system,this paper adopts a DRL algorithm to find an optimal power allocation scheme in downlink two-user NOMA system based on the proposed user pairing scheme.First,the power allocation problem is transformed into a DRL model,and the feature matrix is built according to the signal-to-noise ratio(SINR)and data rate of the two users.Afterward,the asynchronous advantage actor-critic(A3C)algorithm is adopted to find the optimal power allocation scheme.Finally,a joint optimization scheme of the user pairing and power allocation by using DQN-A3C is proposed.Simulation results show that compared with the existing optimization schemes,the proposed scheme can effectively improve the performance of the NOMA system.Finally,this paper makes a detailed summary of the research content of the full text,and prospects for the content which is ready to be studied.