Synchronization Parameters Estimation Of Satellite-based AIS Signals

The synchronization parameters estimation of signals is widely used and has greatapplication values in the area of mobile communication, satellite communication andmeasurement technology. The accuracy of parameters estimation has an impact on theperformance of the signal detection and separation directly. In this paper, the parametersestimation of satellite-based AIS (Automatic Identification System) signals in AWGN(Additive White Gaussian Noise, AWGN) is discussed. The correlation operation, maximumlikelihood and DFT (Discrete Fourier Transform, DFT) are applied to optimize theperformance of the estimation.A frequency-offset estimation method is investigated based on the differentialauto-correlation function. By the means of the high-order processing, code modulationinformation was eliminated. The estimation algorithm that is based on the differentialautocorrelation function and a suitable smoothing function was deduced to estimate thefrequency-offset. Simulation results show that its accuracy is better than the earlier methodsand reduced the computational complexity.The synchronization parameters estimation based on the auto-correlation and DFT isproposed. The algorithm determines the principal value and the expansion of the frequencyand then acquires high-precision estimation by data fitting. Based on that, other parameterscan be estimated by maximum likelihood function. Computer simulations were used toillustrate better estimation accuracy and a large estimation range. The estimation error is lessthan10Hz and the range of frequency estimation is-4800Hz~4800Hz when the signal tonoise ratio is more than1dB, which can meet the requirements of satellite-based AIS receiver.A serial separation algorithm based on the joint timing and frequency estimation anddata-aided synchronization estimation is proposed. The strong signal can be detected based onthe accurate parameters estimation. The symbols and the estimation are used to construct themodulation signal as close as possible to the original signal. The signal separation can beachieved by erasing the modulated signals. It can be seen from the simulation results that theseparation algorithm in this paper has a good performance.