The Research Of Channel Estimation Algorithm For OFDM With DFT And Wavelet De-Noising

With the advancement of science and technology,more and more new technologies have been proposed and used.As the key technology of LTE,OFDM technology has become a hot research for many researchers.The purpose of using OFDM technology is to prevent inter-symbol interference and interference between channels.Channel estimation,a key part of OFDM technology,has been the focus of attention.The channel estimation technique makes the recovery of channel state information becomes more accurate and the performance of the system is improved.The paper focuses on researching the OFDM estimation methods further,and puts forward the corresponding improvement algorithms for the classical estimation methods.The contents and innovation points of the research are described as follows.(1)The basic characteristics of the wireless channel are studied,and the various fading of the wireless channel in the real environment is analyzed.Three kinds of OFDM pilots and three common estimation algorithms are introduced in this paper,which researches the principle of OFDM and its key technologies for the inter-symbol interference and the interference between channels in the wireless channel.Aiming at how to get a good channel response value,some common interpolation algorithms in practical application are researched in this paper.(2)The improved DFT estimation method based on the threshold is studied.After the signal is discretized,it will often cause the spectrum to be leaked.In this case,the Hamming window is proposed to solve the problem.For the classical DFT estimation algorithm can not filter out the problem of noise within the cyclic prefix,a method of adding the improved threshold for de-noising is proposed.The threshold method can effectively eliminate the influence of sudden large pulse noise in the channel,and the improved algorithm is simulated by MATLAB.The experimental results show that the improved algorithm is the smallest in terms of bit error rate and the least in terms of mean square error.Therefore,the improved algorithm can improve the estimation performance well.(3)The improved DFT estimation method based on the wavelet is studied,and which proposed to join the wavelet improved method according to deficiencies of the traditional DFT-MMSE estimation algorithm.Due to the deficiencies of DFT and threshold fusion algorithm,an improved algorithm based on wavelet de-noising and threshold de-noising is proposed in this paper.The algorithm is to add the wavelet decomposition after the LS channel estimation and to remove the noise by using the soft threshold method.Immediately after the discrete Fourier transform of the resulting signal,the threshold processing is performed again to reduce the noise to a lesser extent.The simulation results show that the both proposed algorithms can be a good way to reduce the noise and improve the performance of the system.