The Research Of The Channel Estimation Of OFDM System Based On Pilot-aided

OFDM(Orthogonal Frequency Division Multiplexing) has high utilization of frequency spectrum and good capability of countering multi-path interference. This technology has been regarded as the core technology of 4G wireless communications system in the future. In wireless communications system, the signal needs to be recovered and estimated, so the pulse response of the transmission channel needs to be known. Because the doubly selective peculiarity of the wireless channel, the accurate pulse response of the transmission channel cannot be got, we need to estimate the channel. The method based on pilot get the whole channel response by computing the sub-channel response. Compared with the blind channel estimation, it has lower computational complexity and can track time-varying channel. In this thesis Pilot-aided method was applied to wireless channel estimation.In this dissertation, analysis about doubly selective of wireless channel was given. The model of smooth fading and frequency selective channel was constructed and also provided a simulate method. In the next, the analysis of channel estimation principle based on pilot-aided and impulse response interpolation method was provided. The analysis and simulation to the traditional channel estimation method provided a conclusion that the channel estimation method based on LS has poor accuracy and the method based on LMMES and SVD enhances the accuracy are got, but has a high computational complexity, which need to know the statistical information.For that case, the track property of adaptive algorithm was applied to the estimation of wireless channel. Simulation results have proved the method based on adaptive algorithm does not require any prior knowledge of channel statistics .while guarantee the accuracy of channel estimation, it has low computational complexity. The method based on LMS(Least Mean Square)filter has a property of low computational complexity and the method based on RLS(Recursive Least Square)filter has faster convergence. The adaptive channel estimation algorithm has better MSE and BER performances than traditional methods.