The Research On Interference Management For The Next Generation Wireless Communication Networks

Compared to the existing network mode,the next generation wireless communication network has better performance in transmission rate,system delay and spectrum efficiency.Along with the continuous development and improvement of communication technology,the next generation wireless communication system will be a super-large heterogeneous network(HetNet)with multi-system integration,but at the same time,the problem of interference is more serious,which will seriously affect the data transmission performance and efficiency of the network.Interference management technology aims at eliminating or suppressing the interference caused by spectrum reuse in the network through flexible and efficient algorithms or technologies,and minimizing the loss of system performance caused by interference,which is of great significance to the service performance of wireless communication networks.The interference management technology for the next generation wireless communication is studied in this paper.The main innovative work is as follows:Firstly,a two-stage distributed interference alignment approach based on grouping(GTDIA)is proposed to solve the problem that the types of interference in two-tier HetNet have increased due to the super-intensive deployment of nodes.By dividing the users of each cell into the same group,the algorithm aligns the inter-layer interference(inter-layer IF)signals in each group to the same signal subspace and eliminate them through the precoding matrices at receivers.Based on that,the two-stage transmit precoding matrices are designed and the distributed algorithm that maximize the signal to interference and noise retio(max-SINR)is used to suppress the intra-layer interference between macro users.Simulation results show that the proposed algorithm can effectively improve the system capacity and has a lower algorithm complexity.Secondly,aiming at solving the problem of signal space dimension waste caused by spectrum sharing between the single-hop and multi-hop transmission links in heterogeneous relay network,a cross time slot partial interference alignment algorithm is proposed.The algorithm effectively utilizes the cross time slot characteristics between two-hop transmissions,and aligns part of interference signals contained in the superposed signals transmitted in the first hop to the desired signal subspace of the second-hop transmission.Simulation results show that the proposed algorithm effectively compresses the spatial dimension of the interference signal data streams,and significantly improves the achievable sum degrees of freedom(SDoF)of the system.Finally,a power control scheme based on deep reinforcement learning(DRLPC)is proposed to solve the problem of excessive backhaul link overhead caused by channel state information(CSI)sharing during the interference management.Firstly,the feasibility of using deep reinforcement learning to solve the relay power control in single cell single antenna relay network is veriified.Then,it is extend to the multi-cell multi input multi output cellular network to solve the joint power control problem.In the algorithm,the multi-cell joint power control under time-varying channel is modeled as a Markov decision process(MDP).The central controller is introduced to issue the power control instructions for each base station,and the state space that support learning is constructed based on the strength of the received signal power at each user.The results show that the proposed algorithm can guarantee the sum rate performance of the system to meet the specified threshold requirement,and also effectively improve the frequency spectrum utilization.