Channel Estimation For MIMO-OFDM Systems Under High Mobility Scenes

A new generation of high speed railway mobile communication system LTE-R,with MIMO and OFDM as its core technology,with high data rate,low delay,packet transmission and other characteristics,is one of the most promising systems for future railway communication now.However,LTE-R is facing many challenges,one of the key challenge is how to realize the signal correct transmission in complex and mutable high speed railway scenes.On the one hand,the wireless access system frequently switches between base stations(BS),making the channel environment rapidly and repeatedly change,so the correct transmission of the signal must be based on efficient channel estimation.On the other hand,the Doppler effects are significant,resulting in serious inter carrier interference(ICI),so the correct transmission of the signal must be based on accurate channel estimation.Therefore,this paper focuses on the MIMO-OFDM channel estimation under high mobility scenes,and aims to lay the theoretical foundation for improving the quality of service in high speed railway communication.The main contents are as follows:(1)The MIMO-OFDM channel estimation based on matching pursuit greedy method is studied under high mobility scenes.The regularized orthogonal matching pursuit(ROMP)which is the most widely used now has two drawbacks.First the sparsity should be known beforehand and second once the atom is determined it can not be deleted.According to these two problems,this paper presents an improved ROMP channel estimation method based on compressive sensing by exploiting the sparse structure of channel impulse response.The proposed method combines the advantages of compressive sampling matching pursuit(CoSaMP),sparsity adaptive matching pursuit(SAMP)and variable step size to achieve the reconstruction of sparse signal quickly and accurately.Compared with channel estimations respectively based on orthogonal matching pursuit(OMP),ROMP,CoSaMP and SAMP,simulation results demonstrate that the proposed method effectively improve the performance of the MIMO-OFDM systems in the normalized mean square error(NMSE)and bit error rate(BER)under high mobility scenes.(2)The MIMO-OFDM sparse channel estimation based on basis expansion model(BEM)is studied under high mobility scenes.This paper establishes BEM of the channel impulse response in the delay Doppler domain,and analyzes the sparsity of theestablished BEM.On the basis,the channel estimation problem is transformed into a sparse signal reconstruction problem combined with CS.In order to solve this reconstruction problem,two sparse channel estimation methods are proposed:1)Approximate sparse channel estimation method combined with BEM and pilot pattern(PP-BEM).The method uses approximate processing to the sensing matrix in the solution process.The unknown elements of the sensing matrix are all ignored and modeled as noise and then solved.However,ICI and the reconstruction error of CS are introduced,which reduce the accuracy of channel estimation.In order to improve the situation,an optimal pilot pattern for SISO-OFDM systems and a simplified pilot pattern for MIMO-OFDM systems are designed.Whether in SISO-OFDM or MIMO-OFDM systems,compared with PP-BEM based on continuous pilot pattern or completely random pilot pattern,simulation results demonstrate that PP-BEM based on optimal or simplified pilot pattern performs better in the mean square error(MSE)and BER under high mobility scenes.2)Iterative sparse channel estimation method combined with BEM and improved ROMP(iROMP-BEM).The least squares(LS)method is used to estimate the unknown elements in the sensing matrix to replace the approximate processing of method 1),and reconstructed the estimated value by improved ROMP method.The feedback results are used to gradually improve the accuracy of channel estimation in the iterative step finally.The simulation results demonstrate that the method can effectively improve the channel estimation accuracy,when the number of iterations(t)and ICI bandwidth(D)increases to a certain value(tmax=5,Dmax=2)under high mobility scenes.Furthermore,in MIMO-OFDM systems under high mobility scenes,compared with LS,channel estimation based on CS,PP-BEM combined with simplified pilot pattern and i ROMP-BEM in MSE and BER performance,simulation results demonstrate that the proposed latter two methods can estimate the channel more accurately and effectively,and has stronger robustness in the serious Doppler frequency shift environment.Moreover,iROMP-BEM can further improve the MSE and BER performance of MIMO-OFDM systems,and has higher estimation accuracy.