Research On Compressed Sensing And Interference Suppression Methods Based On MIMO System

The MIMO system with a large number of antennas at the transmitting end and the receiving end has been well used in the fifth generation of mobile communications.Signals are transmitted through multiple antennas at the transmitting end and the receiving end..In practical applications,MIMO-OFDM technology and massive MIMO technology developed from MIMO technology are also widely used in the new generation of wireless communications.Because of its spatial dimension,MIMO technology has also become one of the key technologies for communication.Therefore,the following three aspects of MIMO technology are studied:Firstly,summarize the non-blind estimation,blind estimation and semi-blind estimation in the channel estimation method in wireless communication,and analyze the advantages and disadvantages of several traditional channel estimation algorithms,and then aim at the optimization algorithm proposed in this paper.The PCA-MMSE algorithm is introduced.The principal component analysis of the PCA dimensionality reduction method is performed.The minimum mean square error cost function is being solved "to obtain an optimized channel estimation method based on the PCA-MMSE algorithm.The simulation is performed with traditional classical algorithms.Compared.Through the simulation results,it can be found that the paper uses machine learning algorithms,combines the dimensionality reduction algorithm and the minimum mean square error algorithm,and proposes a new system performance optimization method.The system performance is significantly better than the traditional algorithm.It can improve certain efficiency,reduce pilot overhead,save bandwidth resources to a certain extent,make full use of the frequency spectrum,and balance various parts of the system.Secondly,the principle of the compressed sensing-based reconstruction algorithm is specifically described,and the performance of the channel-dependent sparsity assumption of massive MIMO systems is studied.In many practical situations,the channel is not completely sparse.In addition,in the deep learning-based channel feedback algorithm,the deeper a network is,the less efficient it will be on the training set.In this paper,the attention model in deep learning is used at the encoder and decoder,respectively.The analysis of signal weights improves the performance of channel state information feedback.At the same time,it integrates convolutional neural network knowledge in deep learning and adds residual networks to the system,so that network performance is not affected,which can effectively or even improve effectiveness.Finally,focus on the precoding problem of downlink beamforming in massive MIMO systems.It is considered that in the actual wireless communication environment of a cell,the smart antenna array will be affected due to the limitation of the transmit antenna power,the constraints of the antenna array,and the number of users.In order to reduce the problem of inter-cell signal transmission interference,this paper proposes an optimized beamforming precoding technology processing method.After the beamforming matrix design is initially obtained,singular value decomposition is performed on the high-dimensional precoding matrix to improve the feasibility of the calculation.Secondly,the nature of Lagrangian duality is optimized to reduce complexity and increase Double advantages of system performance.