The Research On The Frequency Offset Estimation In IMT-Advanced System

This thesis introduces the features of two IMT-Advanced international standards of theLTE-Advanced and 802.16m, the basic principles of OFDM in the physical layer core technology,and studies the impact of Doppler shift on the IMT-Advanced system in the high-speed mobileenvironment.With the rapid development of high-speed railway and magnetic levitation trains, Dopplershift caused by the high-speed mobile will seriously affect the performance of IMT-Advancedsystems, thus requiring accurate frequency offset estimation.The commonly used classicalfrequency estimation methods are mainly divided into three categories, this thesis focuses on twocategories based on the cyclic prefix and pilot sequence, gives the specific description of the cyclicprefix-based maximum likelihood algorithm and 5 pilot sequence algorithms, named maximumlikelihood frequency offset estimation algorithm, SCA algorithm, improved SCA algorithm,time-frequency offset estimation algorithm, iterative related frequency offset estimationalgorithm .Although algorithms based on cyclic prefix can improve the spectrum efficiency,it haspoor performance because the length of cyclic prefix is limited, the pilot sequence algorithmestimates based on sequence information, although the increase in complexity, the accuracy ofspectrum estimation has better performance than the algorithm based on cyclic prefix.Classical frequency offset estimation algorithm using a single time slot within a pilotsequence or cyclic prefix sequence caused the larger estimation error by short sequence orinter-symbol interference, even though it can improve the estimation accuracy using two slot pilot,but it can not be used in the situation of two hopping time slot in the IMT-Advanced system frame,so this thesis studies a new single-slot frequency offset estimation algorithm. Do the first frequencyoffset estimation through pilot sequence or cyclic prefix ,then do the first phase compensation to thedata, re-use compensation data for the second frequency offset estimation, improved the accuracy ofthe single slot offset estimation and demodulation performance of the system.Finally, the simulation results verify the improved algorithm's superiority of frequency offsetestimation performance with respect to the classical algorithm.