Research On Key Issues Of User Access And Resource Allocation In Indoor VLC/RF Heterogeneous Network

With the rapid development of mobile Internet technology and various service demands,the spectrum deficit problem is becoming more and more serious.With the advantages of high speed,abundant spectrum resources and good confidentiality,visible light communication(VLC)technology has become one of the effective ways to alleviate the spectrum deficit problem.However,a single communication system consisting of only VLC network has problems such as limited coverage and difficult uplink transmission,so researchers have proposed a VLC/RF heterogeneous network by using VLC network as a base and introducing a radio frequency(RF)network into this network.This heterogeneous network can largely improve the system transmission rate and achieve seamless coverage.This paper investigates the current status of research on indoor VLC/RF heterogeneous networks.Various solutions are proposed for the user access problem of indoor VLC/RF heterogeneous network and the joint optimization of user access and power allocation in indoor VLC/RF heterogeneous network.The performance of these solutions in solving the corresponding problems is also verified by simulation analysis.The main research contents are as follows.First,the system model of indoor VLC/RF heterogeneous network is introduced.The effects of its network model,room size and AP layout on the system performance are analyzed.The transmission model of the heterogeneous network is described.The basis of reinforcement learning and deep reinforcement learning algorithms are described.Next,a multi-factor optimized user access scheme based on deep Q-network(DQN)algorithm is proposed for indoor VLC/RF heterogeneous network user access problem.A multi-factor evaluation model is established to determine the state vector,action vector and reward of the DQN algorithm.The system performance under standard rectangular layout and irregular hybrid circular layout based on different parameters and different environments is simulated and analyzed.Finally,a joint optimization algorithm of user access and power allocation for indoor VLC/RF heterogeneous networks based on D3 QN algorithm is proposed for the problem of joint optimization of user access and power allocation for indoor VLC/RF heterogeneous networks.The differences between the double-depth Q network(DDQN)and duel-double-depth Q network(D3QN)and the DQN algorithm are illustrated.The state vector,action vector and reward of the D3 QN algorithm are determined.The system performance is simulated and analyzed for different parameters and different environments under a standard rectangular layout.