Study On Synchronization Technology For OFDM System

The next generation mobile communication systems aim at providing higher quality and data rate transmission at the minimum lost (power,bandwidth or complexity ) in finite wireless resource. Orthogonal Frequency Division Multiplexing(OFDM)has become one of these key technologies because of its excellent performance in improveing spectral efficiency and its robustness to selective fading channels.Practically, it does not only increase the system's capability, but also can transmit multimedia operation which consists of voice,data and video with higher quality. OFDM is a multi-carrier modulation technique with high performance. In OFDM systems, the whole system bandwidth is divided into several sub-channels with narrow bandwidth and the efficiency of the OFDM systems is increased by allowing overlapping of these sub-channels. Moreover, Cyclic Prefix(CP) which is introduced by OFDM systems avoids inter-symbol interference(ISI) and inter-carrier interference(ICI) effectively. However, if OFDM system can't keep the orthogonality between sub-carriers, all of these advantages won't be guaranteed. Therefore, the OFDM systems require precise synchronization technologies both in time and frequency domain.In this thesis, synchronization algorithms of symbol timing and carriers frequency offset correction for OFDM are analyzed and researched. The research on the non-data-aided synchronization algorithms, which are based on CP, have been mainly concentrate on the classical synchronization algorithm---Maximum Likelihood(ML) algorithm and those algorithms derived from it, such as ensemble correlation algorithm, difference correlation algorithm, the combintion of these two algorithms and the improved algorithm. As for data-aided synchronization algorithm, the Schmidl&Cox algorithm is the first one be introduced. Then Minn algorithm and Park algorithm are introduced, both of them are the improved algorithm of Schmidl&Cox. There is also a new algorithm called CAZAC which has higher accuracy of timing estimation. Each algorithm has its own advantages and disadvantages. So in addition to studying how to improve the algorithm performance, how to using a appropriate algorithm in its best application fields is discussed in this thesis.