Research On Synchronization Technology Of Orthogonal Frequency Division Multiplexing System

television, digital audio broadcast and IEEE 802.11a for its high data speed transmission and high utilization rate of frequency spectrum. It is also one of the key technologies of the next generation of mobile communication system.Because of the reason that the sub-carrier should be orthogonal to each other, the OFDM system is sensitive to the timing error and frequency offset. Timing error would lead to the ISI and frequency offset would destroy the orthogonality of the sub-carriers, all of these would reduce the performance of the system. Therefore, accurate synchronization of symbol timing and carrier frequency is necessary to make the OFDM systems work effectively.Firstly, the research background and research status of the synchronization techniques in OFDM systems are discussed in the thesis, The principle of the OFDM systems and synchronization techniques, the effect caused by the error of timing, frequency offset and sampling clock are also discussed in the thesis.Then, the Maximum Likelihood (ML) algorithm is studied in this thesis, and simulation is also made to analyze its performance. Because of the high complexity of ML algorithm, the Maximum Correlation (MC) algorithm is introduced, which helps to reduce the complexity effectively but will also cause low estimate accuracy. An improved ML algorithm is proposed according to the characteristic of the ML and MC algorithms, and simulation results show that the algorithm can reduce the complexity of the ML algorithm while its performance is better than the MC algorithm.In the end, the Schmidl & Cox algorithm is discussed in the thesis. The Schmidl & Cox algorithm proceed the synchronization by using the training sequence which can estimate both timing error and frequency offset, but its timing accuracy is low. According to this, Minn algorithm and Park algorithm are introduced, which can enhance the accuracy of timing. In the thesis, a new algorithm is proposed which uses only one training sequence to estimate the symbol timing and frequency offset. The simulation results show that the new algorithm will make a accurate timing synchronization, reduce the complexity and use less resource while remaining a great performance of the frequency offset.