Study On Time And Frequency Estimating Technology Of OFDM System

Orthogonal frequency division multiplex (OFDM) is a kind of multi-carrier transmission technique and has been taken as a major technique for the next generation communication system as well as 4G, because it can provide high-speed data transmission in complicated wireless channels with high spectrum efficiency and the abilities in resisting frequency selective fading and inter-symbol interference (ISI). But OFDM is very sensitive to synchronization error, and when existing timing error and frequency offset, it can result inter-carrier interference (ICI) and ISI, and the whole performance of OFDM system will be affected seriously. So studying on high-precise synchronization technique is great important to improving whole performance of OFDM system.At first, this paper overviews the fundamental principle of OFDM technique and meanings of study on synchronization in detail. The whole performance of OFDM system how affected by symbol timing error ad frequency offset is discussed and analyzed over theoretical derivation and simulation results.Then, the timing synchronization algorithms of OFDM have been studied, respectively analyzing two kinds of synchronization algorithms-data assisting and no-data assisting. In this thesis, timing synchronization algorithms including ML based on cyclic prefix,T.Schmidl&D.Cox (S&C),H.Minn and Park algorithms are analyzed in detail and their advantages and disadvantages respectively is analyzing over simulation results.Base on above analysis results, the improved timing algorithm based on Constant Amplitude zero auto-correlation (CAZAC) sequence is proposed, which makes the best of relevant information from CP and train symbol data to estimate timing position. The simulation results show that the improved timing algorithm enhances the precision of timing estimation and effectively overcomes platform phenomenon end much spike phenomenon in timing estimating process, and that phenomenon is resulted from T.Schmidl&D.Cox (S&C), H.Minn and Park algorithms.Last, the frequency synchronization algorithms of OFDM have been studied, respectively analyzing Molrelli&Mengali (M&M) algorithm and frequency offset estimating part of S&C algorithm. Then, at the foundation of timing estimation, improved carrier-frequency synchronization algorithm integrating Moose algorithm is proposed. Simulation results shows that it not only improves the precision of frequency offset estimation but also enlarges the range of frequency offset estimation by single training sequence, as well as improving the performance of frequency synchronization.