Research On Parameter Blind Estimation And Blind Synchronization Of OFDM Signals

Orthogonal Frequency Division Multiplexing technology(OFDM)has the advantages of parallel high-speed transmission,anti-frequency selective fading,and simple channel equalization.At present,it is widely used in various fields,therefore OFDM blind signal processing has become the current research hotspot.The non-cooperative communication party,after capturing the signal,analyzes the received signal using the inherent characteristics of the OFDM signal to obtain the desired parameter and demodulate the OFDM signal.Thus the enemy signals can be listened or even interfered.This paper mainly studies the parameter blind estimation and blind synchronization of OFDM signals.The blind recognition technology and blind equalization technology of OFDM commonly used subcarrier modulation methods are studied,including blind identification system design,modulation blind recognition,parameter Blind estimation,blind synchronization,and blind equalization algorithm design.And simulation analysis is performed.The main contents are as follows:The blind recognition of subcarrier modulation schemes commonly used in OFDM system are studied.Algorithms for estimating the parameters such as bandwidth,signal-to-noise ratio,carrier frequency offset,and symbol rate are discussed,along with the modulation recognition algorithm based on outlet spectrum and constellation matching.The recognition performance is analyzed under the condition of different frequency offset,symbol rate and identification number.The simulation results show that the BPSK,QPSK,16 QAM and 64 QAM signals can achieve more than 90% blind recognition probability when the in-band signal-to-noise ratio is greater than 5dB,6dB,12 dB and 19 d B respectively.Under different parameter conditions,the blind recognition system has better robustness and lay the foundation for blind demodulation of OFDM subcarrier data.Secondly,the blind estimation algorithms of OFDM signal bandwidth,carrier frequency,cyclic prefix length,effective symbol length,OFDM symbol length and modulation symbol period are studied.In terms of bandwidth estimation,an algorithm based on wavelet decomposition is studied.Based on the bandwidth estimation,the center frequency of the extracted bandwidth is taken as the rough estimate of the carrier frequency.In terms of cyclic prefix length estimation,by presetting the proportional scale set,the most suitable cyclic prefix ratio is distinguished by using the kurtosis coefficient.For the blind estimation of other time parameters,the OFDM time parameter estimation algorithm based on cyclic autocorrelation is mainly studied.The simulation results show that the normalized Mean Square Error(NMSE)can reach 410-or less after the in-band signal-to-noise ratio is greater than 3d B,and the error estimation error can be controlled after the in-band SNR is greater than 3dB In a single subcarrier interval.When the in-band SNR is greater than 6dB,the cyclic prefix ratio can reach more than 90% of the estimated correct probability.The time parameter estimation algorithm based on cyclic autocorrelation not only avoids the problem of presetting the cyclic prefix scale set,but also can estimate the time parameters accurately when the in-band SNR is more than 5dB.After that,the OFDM signal time-frequency blind synchronization algorithm is studied.The effects of time-frequency synchronization error on the system are analyzed.The performance of maximum likelihood(ML)algorithm based on cyclic prefix and the algorithm based on cyclic smoothness algorithm in AWGN channel and fading channel are studied.The simulation results show that the timing error of the ML algorithm based on the cyclic prefix is about one sampling point,and the frequency deviation of the single subcarrier is less than 210-in the AWGN channel,when the in-band SNR is greater than 10 dB.However the timing synchronization error based on the cyclic stationary characteristic algorithm is about 8 sampling points.The algorithm based on the cyclic smoothness algorithm is still valid under the fading channel,but the performance of the ML algorithm based on cyclic prefix is seriously degraded.Finally,blind equalization algorithms of OFDM signal are studied.The simulation performance of Constant Modulus Algorithm(CMA)is analyzed under different channels,for which it cannot correct the phase rotation,improved algorithms of Modified Constant Modulus Algorithm(MCMA)and Extended Constant Modulus Algorithm(ECMA)are analyzed.For the multi-mode signal,on the basis of the original algorithms,the Constellation Matching Error(CME)is added,which correct the residual error after the original algorithm is converged.Performances of different algorithms are compared and analyzed under different channels.The simulation results show that the steady-state error of MCMA algorithm is lower,and the convergence rate of ECMA algorithm is faster.After adding CME algorithm,both can converge multi-mode signal.Therefore,the equilibrium point can be found between the convergence rate and the steady-state error to select the optimal blind equalization algorithm.