| With the development of high-speed railway,the number of wireless communication users in high-speed mobile environment is also increasing,which leads to the severe test of high-speed mobile communication services.Orthogonal frequency division multiplexing（OFDM）,as a multi carrier modulation technology,divides the channel in the frequency domain into several orthogonal sub-channels,modulates the high-speed serial signal to the sub-channel for parallel low-speed transmission.It has the advantages of strong anti fading ability and strong anti inter symbol interference ability.However,in the high-speed rail environment,the relative displacement between the transmitter and receiver is accelerated,so that the OFDM subcarriers are no longer strictly orthogonal,which leads to inter subcarrier interference（ICI）.Therefore,it is important to find a channel estimation that better suppresses ICI.This thesis mainly studies the OFDM channel estimation in high-speed rail environment.The research contents are as follows.First,this thesis introduces the research status of wireless communication and the implementation principle of OFDM,analyzes the ICI of OFDM channel in high-speed rail environment,studies the pilot insertion method suitable for high-speed rail environment,and various channel estimation algorithms based on pilot assistance.Moreover,an improved DFT channel estimation algorithm is proposed to solve the problem that the discrete Fourier transform（DFT）interpolation algorithm can not well suppress the noise within the cyclic prefix length in the channel impulse response（CIR）and can not use the effective signal outside the cyclic length.It effectively suppresses noise and ICI by introducing the Kaiser window function,setting the threshold and using the idea of loop iteration.Simulation results show that when the Doppler shift is50Hz,the bit error rate（BER）is 10-2and the number of iterations is 20,the proposed algorithm has a performance gain of 2.5-3db compared with DFT interpolation algorithm,and about 4d B when the number of iterations is 50.Second,in order to improve the estimation performance of OFDM system in high-speed railway environment,a k-means iterative channel estimation algorithm based on approximate complex exponential basis extended model feedback is proposed.Based on the fractional tapped channel approximate complex exponential basis expansion model（FTCA-CE-BEM）,it adopts the principle of minimum mean square error equalization to effectively eliminate the ICI of the received signal.At the same time,the CIR density parameter is introduced,and the improved k-means iterative algorithm is used to eliminate the isolated CIR and better reduce the impact of CIR energy leakage,ICI and noise.Simulation results:When the BER is 10-2and the moving speeds are100km/h and 300km/h respectively,the proposed iterative channel estimation has about 1.5d B and5d B performance gains compared with the central differential Kalman filter algorithm based on the BEM（BEM-CDKF）,respectively.When the signal-to-noise ratio（SNR）is 5d B,the normalized mean square error（NMSE）is 2/3 and 4/5 of BEM-CDKF algorithm respectively.Moreover,no matter in the environment of high or low SNR,the cosine similarity of the proposed algorithm is higher than that of the comparison algorithm.Finally,multiplex input multiple output（MIMO）technology and OFDM are combined for high-speed rail channel estimation.Firstly,the MIMO-OFDM pilot is designed,and the optimal pilot design method of inserting the guard pilot and minimizing the mutual incoherence characteristic（MIP）by cyclic iteration is proposed.After inserting the pilot,the BEM algorithm is used to obtain the channel response between each pair of transmitting and receiving antennas,and then detected by MIMO to obtain the estimated value of the transmission signal and the feedback signal filtered by ICI on each pair of antennas.Finally,the improved k-means algorithm is used to filter the noise and ICI,and the channel response of each pair of transceiver antennas is obtained.Simulation results:When the BER is 10-2and the moving speed is 100km/h and 300km/h respectively,the proposed algorithm has about 4d B and 1.7d B performance gain compared with the BEM-CDKF algorithm.When the signal-to-noise ratio（SNR）is 5d B,its NMSE is 7/9 and 5/7 of BEM-CDKF algorithm respectively.It shows that MIMO-OFDM system is suitable for high-speed railway information transmission,and its performance is better than OFDM system. |
