Research On The Techniques Of Synchronization And Channel Estimation In OFDM Systems

Orthogonal Frequency-Division Multiplexing (OFDM) has high spectrum efficiency and is robust to the multi-path fading. Owing to these advantages, OFDM has been used in the European digital broadcast radio system and been selected as the standard physical layers in IEEE802.11a, HiperLAN/2 and IEEE802.16 WLANs. Whereas, some techniques in OFDM systems are still not consummate, especially in Synchronization and Channel Estimation.The first research work of this thesis is Synchronization. In the first place, the influence of carrier frequency, symbol-timing and sampling-timing offset has been discussed. The conclusion is that symbol-timing and frequency offset are the main factors which affect the system performance. In the second place, Maximum Likelihood (ML) method which is based on the Cyclic Prefix (CP) has been researched in-depth. Last, in the data-aided algorithms, Schmidl, Minn and Park methods are studied and simulated by MATLAB.The second analysis point in the thesis is Channel Estimation. This thesis focuses on the pilot symbol-aided estimation. First, I investigate the performance of various pilot arrangement patterns. Second, Least Square (LS) and Minimum Mean Square Error (MMSE) algorithms have been studied for the purpose of obtaining the channel information at the pilot spots. Moreover, some interpolation algorithms have been analyzed to estimate channel information beside pilot spots, such as Linear and Gaussian interpolation. Last, simulations have been carried out to contrast different pilot patterns in various channels.