The Study Of Frequency Offset Estimation Algorithms For OFDM Systems

Orthogonal frequency division multiplexing (OFDM) is a kind of multi-carrier modulation in wireless environments. In recent years, as the rapid development of the hardware technique, OFDM has been given more attention because of its high efficiency of using the limited spectrum and its robustness against the effects of multi-path propagation which are main obstruction for mobile reception and realization of single frequency network.Presently, OFDM has been employed in different areas. It has been adopted as the new European digital audio broadcasting (DAB) standard, as well as for the terrestrial digital video broadcasting (DVB) system. In 1999, IEEE also adopted OFDM as the Physical layer's modulation standard in IEEE 802.11a.While in the meantime, OFDM has its own shortcomings: its sensitivity to a frequency offset and the high peak-to-average ratio. How to eliminate the two problems is the main task of the research of OFDM. In this thesis, we mainly discuss the problem of frequency offset estimation.In this paper, we first introduce the characteristic of wireless channel and the basic theory of OFDM; then we present the reason of bring the frequency offset, the impact on the OFDM system and the losing of signal-noise-ratio. We also look back the existent literatures about the OFDM frequency offset estimation. In the fourth part of this thesis, we analyze the pilot-assisted frequency estimation algorithm based on the use of the maximum likelihood estimate. This method can reach good results in AWGN channels. But it is another thing in multi-path channels, because of the different signal-noise-ratio in different position of the pilot. Then we present a new pilot-assisted frequency offset estimate algorithm based on the use of Weighting Function. In particular, the frequency offsetsestimated from pilots in different spectral locations are properly combined together so as to reduce the influence of multi-path channels. Simulations show that the performance of the frequency estimator based on Weighting Function provides an excellent estimate of frequency offset in wireless systems.