The Application Research Of Time-varying Channel Estimation Based On FrFT

Fractional Fourier transform can deal with nan-stationary siginals well ,which contains the information of time domain and frequency domain in the fractional domain.It is a local transformation.With the rapid development of the high-speed railway and expressway,people are facing more and more complicated electromagnetic envionment.Channel time-varying characteristics are more prominent, if effective channel estimation at the receiving and isn't gol detalled channel informatin will not be got, which will scriously affect the performance of the whole communieation system.In order to satisfy the high qualsty communicaion requirements,it is, of great value to study the arameters estimation of time-varying channels.The main works of this paper are as follows:1.Duc to the difference of the estimated values and theoretical values for the estimation of signal parameter using the dlecoinposed FrFT fas algorithm.this paper researches dsmensional normalization ,and give how to chooes dimensional normalization factor,Due to energy leakage and fence effect caused by urnucation phenomenon in time domain,the spectrum corrcction technology is analysed; The simulation verifies thai the FrFT of Gauss signal in fractional domain Btillis Gauss sigiul, the FrFT of show natrut of the best properties ofenergy accunulation in the optimal frational transform angle. The selection of the dimension normalization factor is related to the observation time and bandwidth of the seignal.If the observation time and bandwidth are different,the maximum values of the signal in the axis of p and u is diferent.2.If chirp rates of LFM signal used to detect channel are known,the method of channelestmation based on the time shift and frequency shift of FrFT is sudied.In order to improve the speed of searching peak ,he dichotony is used to search the spctrum peaks of LFM signals in fractional domain.As far as possible to reduce the fence effect and the enegy leakage,spectral correction technique is used to correct specral peak.Simulation analysese show the effectiveness of the proposed method.3.If chirp rates of LFM signal used to detect channel are unknown,the method based on RAT and AMSFDF used to estimate channel time delay in the case of existing time delay only;Then the method based on RAT and AMSFDF of time-varying channel parameters estimation is researched.At last the method based on RAT and FrFT of time-varying channel parameters estimation is researched.Binary search used to search the peak of signals combining spectrum correction technique is used.Simulations show the effectiveness of the proposed method. The mean square errors of the proposed time-varying channel parameters estimation methods are reduced and the computation speed is faster.Satisfactory results can be obtained even if the LFM used detected channel frequency modulation rates are unknown ,too.