Carrier Synchronization Technology For Burst-Mode Signal

In the digital communication system, accurate frequency estimation of received signal is a very classical subject in carrier synchronization technology, the existence of frequency offset directly impact on the performance of the receiving system, accurate frequency offset estimation and compensation is very important for demodulation, therefore, a detailed study of frequency estimation algorithm has an important value.This paper focuses on the study of frequency estimation algorithms for M-ary Phase-Shift Keying (MPSK) signal and Quadrature Amplitude Modulation (QAM) signal, and its key point lies in two parts: In the first part, the preliminary frequency estimation methods of the digital communication signal are discussed, which mainly based on the methods of modern spectral estimation. Firstly, briefly introduce the modified Pisarenko Harmonic Decomposition method, which is got by replacing the sample covariance in Pisarenko Harmonic Decomposition with the help of MC frequency estimation; Then, AR spectral estimation method based on the modified covariance algorithm, and the performance comparison of the two algorithms are presented, results show that the performance of AR spectral estimation method is better than the modified Pisarenko Harmonic Decomposition method.In the second part, we discussed the data-aided and non-data-aided frequency offset estimation: Firstly, descriptions of Kay, Fitz, and Luise&Reggiannini data-aided frequency offset estimation algorithms are provided; Then, discusses a non-data-aided feedforward frequency offset estimation; At last, a algorithm which can get accurate frequency estimation in low-SNR condition is proposed, first, the preliminary estimation values can be obtained using AR spectral estimation method based on the modified covariance algorithm, after frequency compensation, the needed frequency offset values can be obtained by combining the M-th transformation and the maximum likelihood criterion, also performance comparison with feedforward frequency offset estimation is presented, simulations results show that the algorithm can obtain needed steady precision under low-SNR condition. All algorithms in this paper are given Monte Carlo simulation results and compared with the Modified Cramer-Rao Bound.