The Research And Simulation On OFDM Synchronization Algorithm

Orthogonal frequency division multiplexing (OFDM) technology is a multi-carrierdigital modulation technique. It has advantages of strong anti-multipath capability and highspectrum utilization rate, which applied to the fourth generation mobile communicationscombined with other techniques. But OFDM technology is very sensitive to timingsynchronization error and frequency deviation. Timing offset will cause the rotation of thephase and frequency offset makes the orthogonality of each OFDM sub-carrier destructed,which has a sharp deterioration in OFDM system performance.This dissertation researches OFDM synchronization algorithm and simulation, optimizessymbol timing synchronization and frequency synchronization algorithm as the mainproblems. Firstly, it introduces the outcome of OFDM synchronization technology at homeand abroad, and a detailed analysis of symbol timing error and carrier frequency offset ofOFDM system. We analyze their own advantages and disadvantages of timingsynchronization methods based on cyclic prefix and training symbol and show intuitivelywhether the timing performance is good or bad by simulation diagram.Secondly, the dissertation points the shortcoming that the muti-correlation peak of timingestimation curve causes the timing error owing to the cyclic prefix in OFDM preamble-aidedsynchronization algorithms, and proposes a improved algorithm. The improved proposal isthat redesign the training symbol by introducing the PN weighted factor and propose thecorresponding timing metric using the autocorrelation of the PN and the conjugatedsymmetrical structure of the advanced training symbol to achieve timing and frequencysynchronization. After the timing synchronization and frequency correction, BER of theOFDM transceiver system is simulated. The error rate declined is proof validly.Thirdly, to overcome the more similar strength correlation peak with the correct timingposition because of the different length of cyclic prefix of the CAZAC sequence, Timingalgorithm eliminates the multiple correlation peaks by adjusting the structure of thepseudo-random sequence. Timing estimation performance of symbol synchronizationalgorithm based CAZAC sequence is improved.