The Research Of The Timing And Carrier Frequency Synchronization Algorithms For OFDM Systems In Wireless Channel

OFDM is an effective data transmission technology which can be against multi-path propagation, inter-code interference and narrow band interference, and it also has a high spectral efficiency. But it is more sensitive to synchronization errors, which bottlenecks the development of OFDM technology. The algorithms of time and frequency synchronization for OFDM systems are mainly investigated in this dissertation.Above all, this dissertation gives an overview on the development of OFDM technology. Transmission principles and system parameters design are introduced. Based on the basic characters of wireless channel, the mathematics model used for simulation is gotten. The effects of timing and frequency synchronization errors on the performance of OFDM system are systemically summarized in this dissertation, and the requirement precision of time and frequency synchronization is given. In the time synchronization, the algorithms based on three different PN sequences structures and pilot phase detect are discussed, and the figure of time emendation is proposed at last. In frequency synchronization, the algorithm of integer frequency offset estimation based on pilot is discussed, then some classical frequency offset estimation algorithms is given, and this dissertation points out the drawbacks of them. A new scheme based on cyclic-shifted preamble sequences is proposed for frequency synchronization in OFDM system. According to the time property of DFT, the preamble sequences are constructed and two estimators are designed. The coarse frequency offset is realized by the first estimator. Once coarse frequency offset compensated, the fine estimation is achieved by the optimized line combination of two estimators. In multi-path channel, two estimators are optimized and can eliminate the floor phenomena. It is demonstrated that, compared with the classical algorithm, the new method has better estimation accuracy and larger estimation span, which can be changed more flexible. At the same time, the Cramer-Rao lower bound of the new method is analyzed and obtained. Finally, this dissertation gives three non data-aided algorithms theory deduction and simulation performance respectively, and analyses their advantages and shortcomings.