Research On Channel Estimation Algorithm Of LTE Downlink

LTE-Advanced system is currently recognized as the fourth generation mobile communication system. LTE-Advanced system has faster data transmission rate, larger system capacity and higher spectral efficiency of the system with peak rate has been significantly increased. The critical technologies adopted in LTE-Advanced system are mainly OFDM and MIMO.Under high-speed mobile environment, channel of LTE Advanced system changes quickly in a short period of time, so it’s necessary implementing accurate channel estimation in order to precisely restore signal at the receiving end.First of all, the current research status of channel estimation technology in the domestic and overseas are analyzed in this thesis, then the physical channel structure, baseband signal processing procedure and key technologies of LTE downlink are introduced. On the basis of analyzing the fading channel characteristics of high-speed mobile scenario, LTE downlink system channel estimation simulation platform is established.Secondly, BEM fitting models are studied. In addition, CE-BEM, DCT-BEM, P-BEM, DKL-BEM, DPS-BEM and GCE-BEM combined with oversampling technology are briefly introduced, and performances of all these fitting models are simulated and analyzed, meanwhile, their mean square error are compared.Subsequently, performances of pilot LS, LMMSE, SVD-MMSE and simplified SVD channel estimation algorithm are analyzed on LTE downlink link simulation platform including different send and receive antennas, different MIMO detection algorithms, different modulation modes, different scenarios and different terminal speed. As a result, it is concluded that the performance of LS algorithm is the worst, while the performance of LMMSE and simplified SVD are the best.Finally, BEM fitting models are applied to LTE downlink system channel estimation. Performances of channel estimation under high speed mobile environment are studied in the thesis. It is shows that BEM models could well match the real channel and channel estimation performance is better than that of LS algorithm. Particularly the change of BEM channel estimation performances are analyzed under the scenario in which the terminal speed is120km/h or350km/h, and it is concluded that the changing of terminal speed influence a lot of estimation performances.