Research On Several Problems Of Interference Management In Broadband Mobile Communications

With the rapid development of society and economy,the data service of mobile users has been exploded.To realize the demands and vision for low delay,high connection number,high density and high mobility,we will introduce some new technologies such as ultra-dense networks(UDN),full duplex(FD),D2D(Device-to-Device)communications,etc.The introduction of these advanced technologies could increase system capacity and spectrum utilization,but also brought some serious and complex interference problems.How to effectively eliminate self-interference and reduce inter-user and inter-cell interference are key technical issues in broadband mobile communications.In this paper,the interference management problems of user self-interference,inter-user and inter-cell interference are studied,and the local-to-global optimized interference management are carried out.Fist,for the problem of self-interference that the FD transceiver will be interfered by its own signal when receive interferce from When transmits and receives signals at the same frequency simultaneously.We study the self-interference cancellation technology in the full-duplex system.The spectral estimation algorithm is used to linearize the in-band self-interference channel.The genetic algorithm is used to optimize the parameters of the adaptive filter.The improved algorithm based on Least Mean Square(LMS)filtering is designed and implemented.Simulation results show that the improved algorithm proposed in this paper has a stronger anti-interference ability and better convergence.Then,for the problem of massive D2D users simultaneously handover on cell edge with high-speed will lead to congestion,the interference coordination technology in the D2D communications is deeply studied.We propose a speed-aware joint handover approach that leverages D2D multicast and D2D clusters.The proposed solution can effectively alleviate the pressure of broadcast signalling storm,and improve the handover QoS in terms of interruption probability and handover latency.In addition,we attempt to address the challenge of a large D2D UE population in mode selection.We employ a gametheoretic approach and formulate an evolutionary game,based on which a distributed device-controlled algorithm is designed for D2D mode selection.Using three mode for uses to selecte which are cellular mode,direct reuse mode and relay mode.The simulation results show that the proposed scheme can achieve the highest utilities compared to the max-SINR scheme,the distance based scheme and the random scheme.At the last,for the problem of different layer interference management of complex network structure,the research on interference avoidance technology under ultra-dense networks is carried out.On the one hand,we construct a system model under ultra-dense networking.Then according to the complex resource allocation problem in the networking,a resource allocation algorithm based on conflict factor graph is proposed.The resource conflict graph is constructed by using resource conflicts,reduced the complexity by using the factor graph,and solved the multi-objective problem of resource allocation and power allocation.The simulation results show that the proposed algorithm not only improves the edge throughput,but also improves the average throughput of the whole network,and improves the spectrum efficiency while reduce the interference between users.On the other hand,an reinforcement learning power allocation algorithm based on graph signal processing is proposed.The graph signal processing tool is used to analyze the network interference,and the interference analysis result of the whole network is obtained as the state to optimal reinforcement learning power allocation.The simulation results show that the proposed algorithm improves the throughput of edge users while ensuring the peak rate,thus improving the throughput of the whole network.