Study On Channel Estimation Techniques For The Improved OFDM Systems

Future communications will face a wide range of application scenarios,thus requiring the new air-port waveform techniques exhibit more attractive features,such as low latency,allowing dynamic spectrum accessing,good expansibility,etc.The technique of Orthogonal Frequency Division Multiplexing(OFDM)as the signal transmission scheme of 4G mobile communications downlink,due to its some native drawbacks such as high out-of-band radiation and peak to average power ratio,sensitivity to frequency offset,single waveform parameter in full spectrum band,etc,can't handle the communications of future various application scenarios completely.The OFDM systems based on Fractional Fourier Transform(FrFT-OFDM)as one of new multi-carrier modulation technique,uses the chirp basis function to replace the sinusoidal basis function.That exploits Fractional Fourier Transform(FrFT)to modulate and demodulate,so sub-carrier frequency is not a fixed value but linear with time.The FrFT-OFDM systems can choose the optimal transform order to obtain a best performance,improving the ability to combat the time selective fading.In this sense,the traditional OFDM is a special case when the order of FrFT-OFDM is equal to 1,so FrFT-OFDM is a stretched form of OFDM.In this dissertation,based on FrFT-OFDM,the PAPR performance of FrFT-OFDM and channel estimation techniques are investigated.The ideas of traditional OFDM system channel estimation algorithms such as Least Square(LS),Linear Minimum Mean Error(LMMSE),and Singular Value Decomposition(SVD)are expanded to the system and channel estimation in Fractional Fourier domain is proposed.Than the feasibility and applicable conditions of those algorithms are verified througth a series of emulation experiments.Then referenced the Generalized Frequency Division multiplexing(GFDM)technique theories,the Generalized Fractional Division Multiplexing(GFrDM)which is multi-parameter adjustable is proposed,and the PAPR and channel estimation techniques of the system are investigated.The research content of this paper is summarized as follows.In the first part of the dissertation,according to the principles of fractional Fourier transform,the actual algorithm of Discrete Fractional Fourier Transform(DFrFT)for the modulation and demodulation of chirp subcarriers is theoretically derived.Than combined with theories of multi-carrier systems,the mathematical models of FrFT-OFDM systems are established and the way to add the chirp cyclic prefix is introduced in detail.The baseband models of the FrFT-OFDM systems are built respectively by Matlab.Than the BER performance of the systems under Gaussian white noise and the system parameters affecting PAPR are analyzed through some simulation experiments.Secondly,the wireless transmission channel models are set up to study the channel estimation techniques at the receiver side of the FrFT-OFDM systems.The pilot arrangement,the estimation algorithm of channel fading values and the interpolation algorithms are introduced.The influence factors to the performance of the estimating algorithms is analyzed.The performance of the three algorithms in different selective fading channels are compared.The feasibility of these algorithms are verified.Finally,on the basis of FrFT-OFDM,the Generalized Fractional Division Multiplexing(GFrDM)as a more general technique is proposed.So the mathematical models of the GFrDM systems are derived theoretically.The BER performance of those systems under Gaussian white noise and the systems parameters that affectes PAPR are analyzed through simulation experiments.A new estimation algorithm for GFrDM system that comb-pilotbased fractional-domain LS channel is proposed,and the validity of the algorithm under different selective fading channels is verified.