Research On Key Technologies Of Precoding In Interference Alignment Networks

The data traffic of the future mobile communication system will reach more than a thousand times today.In order to meet such a high demand,the 5th generation mobile communication system(5G)will achieve intensive deployment.This will cause more serious interference problems.How to effectively manage and utilize interference becomes a major research hotspot in the future of mobile communications.Interference alignment(IA)technology can effectively eliminate inter-cell interference and increase system capacity.It is considered to be a very potential network interference management technology.The interference alignment pre-codes the data at the transmitting end so that the interference signals will be aligned to a sub-space which is different from the desired signal at the receiving end,and then the interference can be eliminated through decoding.The interference alignment technology compresses the space occupied by the interference signals,thereby increasing the spatial freedom of the desired signal and achieving the purpose of eliminating interference and increasing network capacity.The key to interference alignment technology is the design of the precoding matrix,while the precoding matrix completely depends on the global channel state information(CSI).In the frequency division duplex system,since the uplink and downlink channels are not reciprocal,accurate acquisition of global channel state information requires the base stations to send pilots and users to estimate and feedback channel information,which causes high system overhead.Therefore,the contradiction between the overhead of channel information feedback and the accuracy of channel information acquisition is unavoidable.In this paper,we study the contradiction.Firstly,we quantitatively analyze the relationship between channel state information acquisition overhead and interference alignment performance.Then we propose a cycle-adaptiveadjusted finite feedback scheme.Based on this,we design and develop a hardware implementation scheme for interference alignment precoding based on limited feedback.The specific research contents of this paper are as follows:Firstly,this paper introduces interference alignment techniques in a three-cell scenario.It describes the principle of interference alignment techniques and the closed-form solution calculation and decoding matrix steps.This paper briefly analyzes the main reasons for limited channel state information at the transmitter and summarizes the current limited feedback schemes for several channel information in multiple-input multiple-output(MIMO) systems,providing a theoretical basis for subsequent chapters.Secondly,this paper analyzes the impact of CSI quantization error,channel matrix compression and feedback delay on feedback quantity and interference alignment performance,and analyzes the time domain correlation and frequency domain correlation of the channel.On this basis,this paper proposes an interference alignment algorithm that adaptively adjusts the channel information feedback period.Using the time-dependent characteristics of the channel,through the channel prediction technology and the joint regulation of the transmitter and the receiver,the length of the feedback cycle is adaptively adjusted according to the channel change characteristics.Then the performance advantages of the algorithm are verified through simulation.Thirdly,this paper studies the application of interference alignment algorithm in 5G massive MIMO system.A scheme combining the interference alignment technology with the joint space division multiplexing technique is proposed to solve the interference problem in a multi-cell massive MIMO network.This scheme eliminates inter-cluster interference through beamforming,and then uses interference alignment to eliminate intra-cluster interference,thereby increasing the system transmission rate.Finally,this paper designs a hardware implementation scheme for interference alignment precoding calculation based on limited feedback.In order to save hardware resources,a lowcomplexity codec matrix calculation algorithm is designed and implemented in FPGA.Then the platform is verified in the actual physical channel environment.