Research On Channel State Information Acquisition Of Massive MIMO System

Massive MIMO technology significantly improves the system spectrum efficiency and user efficiency,and enhances the reliability of transmission.For massive MIMO systems to give full play to their potential advantages,it is crucial to obtain accurate channel state information(CSI).Channel estimation at the receiver to obtain channel state information at the receiver is a key step for massive MIMO systems to obtain CSI.However,in massive MIMO systems,due to the large signal dimension,the traditional channel estimation method will generate huge pilot cost,especially in massive MIMO systems under FDD mode.Compressed sensing(CS)technology can effectively save pilot cost.In cs-based channel estimation,pilot design and reconstruction algorithm are closely related to estimation performance,which jointly determine the reliability and effectiveness of channel estimation.Therefore,this thesis will carry out the research on channel estimation of FDD massive MIMO system from these two aspects.1.Combined with CS,the channel estimation problem of FDD massive MIMO system is modeled as the sparse signal reconstruction problem in CS.According to the minimum of the cross-correlation value of the sensing matrix,a pilot design algorithm based on differential evolution algorithm is proposed.After continuous evolution,mutation,crossover and selection,the optimal location set of pilot sequences is obtained.Simulation results show that the pilot design algorithm based on differential evolution can effectively reduce pilot cost and achieve better channel estimation performance.2.On the basis of the above research,the spatial and temporal correlation of massive MIMO channel is further analyzed,and the structured compressed sensing technology is applied to the channel estimation of massive MIMO system.A structured compressed sensing reconstruction algorithm based on space-time correlation is proposed in this thesis.This algorithm can adapt to the channel sparsity and adjust the step size according to different iteration stages,which solves the problem of slow step growth and overestimation of traditional reconstruction algorithm.Moreover,it is not necessary to take the channel sparsity as the prior information,and further improve the accuracy and efficiency of the reconstruction algorithm by taking advantage of the channel sparsity characteristics.At last,how to group antennas reasonably to improve the effectiveness of channel estimation is given.Simulation results show that the performance of the proposed algorithm is significantly better than OMP,SP,StOMP and SAMP,which significantly improves the channel estimation performance of massive MIMO systems.