Research On Channel Estimation Technique Of Passive Irs Assisted Millimeter Wave Massive MIMO System

Millimeter wave can bring great capacity and performance improvement for wireless communication,but the communication system that integrates millimeter wave also faces the problem of low system performance due to propagation characteristics.Multiple Input Multiple Output(MIMO)technology and Intelligent Reflecting Surface(IRS)technology came into being to make up for the deficiency of millimeter wave signal propagation characteristics.The IRS technology can independently control the amplitude and phase of the reflected signal through the surface reflector,so as to adjust the beam direction to provide beam gain and suppress noise.For purpose of achieve active and passive beamforming in IRS auxiliary systems,a high-precision channel estimation scheme is essential.In this paper,the millimeterwave large-scale MIMO communication system is taken as the background,in order to solve the channel estimation problem in the IRS fusion system,the single IRS case is studied,and then the double IRS case is further studied.The proposed algorithm and scheme can better reduce the training cost and obtain accurate Channel State Information(CSI).On the one hand,in the millimeter-wave large-scale MIMO scenario assisted by single IRS,Parallel Factor(PARAFAC)tensor model is introduced to solve the problem that passive IRS cannot estimate a single channel,and the signal in the system is modeled into a third-order tensor form.In view of the problems of high computational complexity and slow convergence of existing algorithms,this paper proposes a fast hierarchical block iterative algorithm based on Angle subspace projection for multi-dimensional matrix fitting,reducing the size of the tensor.Simulation results show that the proposed algorithm can converge quickly and achieve higher accuracy with less pilot frequency consumption.On the other hand,combined with the idea of compressed sensing and tensor fitting,the research is extended to double IRS assisted millimeter wave large-scale MIMO scenarios.Firstly,to solve the problem that CSI cannot be obtained separately due to the coupling of single reflection cascade signal and double reflection cascade signal of single user passive IRS,a three-stage single user channel estimation scheme is proposed in this paper.By analyzing sparsity of single reflection cascade channel,The channel estimation problem is transformed into Multiple Measurement Vectors(MMV)problem,so as to realize accurate channel estimation by using its structured sparsity.The decoupling signal is obtained by using the known single-reflector cascade channel results and modeled into PARATUCK model.On this basis,an alternate iteration fitting algorithm is proposed,which can get better estimation performance than the classical algorithm.Secondly,the joint scaling property of the common channel is analyzed to solve the problem of the high pilot cost in the case of multiple users.The multi-user channel estimation is carried out based on the single user estimation results.The simulation results show that the proposed method has good estimation performance and reduces the pilot consumption.The channel estimation optimization algorithm and design scheme for single IRS assisted and double IRS assisted wireless communication system proposed in this paper can be well applied to urban or indoor scenes,and provides a research idea for obtaining channel state information in IRS-assisted millimeter wave large-scale MIMO communication system,which has important theoretical significance and application value.