Study Of Synchronization Technology For Ofdm Systems

Orthogonal frequency division multiplexing (OFDM) is a kind of multi-carrier modulation technique, which has advantage of high spectrum efficiency and robust transmission against multi-path interference. OFDM has been widely used in various fields and will be a key technology for next generation of mobile communication systems. However, due to the nature of multi-carrier modulation technology, OFDM is very sensitive to synchronization errors. How to realize the synchronization accurately has become one of the most important matters in OFDM. Therefore, time synchronization and frequency synchronization have become the key issues in OFDM system and a research hotspot at home and abroad.The training sequence synchronization algorithms including the synchronization algorithms based on the PN sequence and the synchronization algorithms based on CHU sequence are studied deeply in this paper. In order to solve the problem of the "peak platform" in Schmidl&Cox (S&C) algorithm which is one of the synchronization algorithms based on PN sequence, an improved timing synchronization algorithm is proposed in this paper, in which the training sequences are weighted by pseudo-random training sequence (PN). The improved algorithm can eliminate the "peak platform" in the timing function curves and improve timing accuracy. By studying synchronization algorithm based on CHU sequence, it has been found that the traditional cross-correlation timing synchronization algorithms are susceptible to frequency offset. To solve this problem, another improved synchronization scheme which is based on the novel structure of training symbol and timing metric function is proposed. By theoretical analysis, this optimized design can overcome the effects of frequency offset, and the accurate timing can be obtained. The two improved methods above are simulated in the Gaussian channel and multi-path fading channel simulation environment, respectively. The simulation results show that the two proposed synchronization algorithms both can achieve better timing performance compared to the corresponding conventional timing synchronization algorithm.