Research On Synchronization Algorithms In OFDM System

Orthogonal Frequency Division Multiplexing (OFDM) is the key technique of the fourth generation mobile communication. The advantage of OFDM is to resist of the multi-path channel. OFDM improves the utilization efficiency of the spectrum. It enhances the capacity of the system and meets the request of the high speed of data transmission and user's movement for the next generation mobile communication.OFDM is more sensitive to the frequency offset and the phase noise than the single carrier communication system, since the frequency offset will break the orthogonality of the sub-carriers, introduce the interference of the sub-carriers, and bring great influence to the system performance. Synchronization is one of the key techniques of the OFDM. This paper . chiefly studies the synchronization of OFDM, and includes the algorithm and actualization.In this thesis, the OFDM model and its principle are introduced firstly in brief, then discuss strengthens and weakens of OFDM system, further more is the synchronization issues, especially focuses on synchronization issues: Analysis the effects of demodulation performance, including frequency offset and timing offset, and in-depth research on symbol timing synchronization and frequency synchronization of OFDM system. A lot of computer simulations are given over AWGN and frequency selective fading channels. We analyze and compare the performance of a few synchronization methods, including their applying areas, merits and shortcomings. Furthermore, the article proposes an improved OFDM timing synchronization algorithm and frequency synchronization algorithm. The improved algorithm could find the exactly timing point and frequency offset by the character of correlation, the simulations show that this algorithm has an obviously peak at exact point and significantly reduces the mean square error of timing estimate. Using the result of timing and reset coefficient to calculate frequency offset. Simulations show that improved frequency algorithm could reach smaller MSE.