Research On Power Allocation Technology For Beyond 5G Multi-service Heterogeneous Networks

In order to meet the flexible requirements of future wireless communication systems for supporting multiple services,the flexible duplex technology is proposed to enhance system performance,which is particularly important in heterogeneous networks enabling vertical industries.When two adjacent cells based on the dynamic time division duplexing technology have opposite transmission directions and share the same time-frequency resources,serious inter-cell interference may occur,which is called cross link interference(CLI).An important scenario for 5G to empower vertical industries is the industrial internet of things,where simultaneous transmissions of many different types of services,such as voice,data and video on different transmission directions becomes normal.However,this also leads to complex interference problems.In this case,how to further optimize the wireless resource management algorithm and effectively alleviate the negative impact of interference is an important problem that needs to be studied urgently.Firstly,this paper introduces the background and importance of studying multi-service power allocation technology for beyond 5G in heterogeneous networks,analyzes the problems existing in heterogeneous networks,and introduces the research progress of power allocation for multiple services at home and abroad.Based on the above investigation,we propose a new heterogeneous network system model,analyzes the existing interference,and allocates the transmission power of the downlink multi-antenna macro base station and the uplink single-antenna small-cell users with the optimization objective function of maximizing the sum of uplink rate and downlink rate.A heterogeneous network power allocation based on improved intelligent algorithm is proposed,which alleviates the limitations of particle swarm optimization and simulated annealing in searching for the optimal solutions.Simulation results show that the proposed algorithm can avoid falling into the local optimal solution,further improve the global search ability,and realize the effectiveness and efficiency of power allocation.Next,by adding multi-service requirements on the above research,we propose a family of directed Q-Learning based power allocation methods for uplink/downlink multi-service concurrency,where the mean opinion score(MOS)is invoked as a metric to characterize users’ quality of experience(QoE)for multiple services and to assist the transmission power allocation on the base station and the users.By improving the update mode of the Q-table after new user join the system,a family of directed QLearning algorithms are proposed.Simulation results show that when the number of users changes,the improved algorithms maintain reasonable MOS values and congestion rate,while reducing the number of iterations and achieving improved convergence performance.