Research On Signal Processing Technology Of AIS Receiver

Automatic identification system(AIS)plays a great role in ship collision avoidance and maritime supervision by exchanging data in the very high frequency(VHF)band.At present,the research of AIS is relatively slow in China,and the core technology is still controlled by other countries.Although some relevant equipment of AIS meets the requirement of the protocol specification,they still can be improved in terms of performance.Therefore,the research of signal processing technology of AIS system receiver is of great significance and application value to promote the localization of AIS system and improve the performance of domestic AIS equipment.The main research contents of this thesis include the following aspects:First,the system of AIS based on the international protocol is introduced in this thesis.To begin with,the architecture construction of the AIS signal processing system is completed,and the message frame structure and various encoding and decoding methods involved in AIS transmission are explained.Besides,the baseband signal used in the AIS system is analyzed,that is,the gaussian filtered minimum shift keying(GMSK)modulated signal.Focusing on its modulation method,the quadrature modulation method of waveform storage is selected as the transmission signal generation method.What’s more,the AIS receiver signal processing flow is given according to the receiver architecture,which lays the foundation for the subsequent design of specific solutions.Second,the receiver architecture,technical indicators and baseband signal characteristics are combined to complete the comparative design of signal detection and signal synchronization schemes.And the parallel processing mechanism of burst signal detection and frequency estimation is optimized to shorten system waiting time,reduce data storage,and reduce Hardware scale.Initially,aiming at the unique burst communication mode of the AIS system,the burst signal detection is completed in the passband by using the sliding window frequency domain detection.Then,in order to counter the ±3ppm limit frequency deviation required by the system index,a shift-addition second-order cyclic cumulant method is proposed.This scheme is less dependent on the signal pattern and the frequency estimation is more accurate.But the corresponding sampling frequency is high,the data cache is large,the system time is extended,and the hardware scale is large.In addition,the square delay method is adopted and the algorithm parameters are determined as a symbol timing synchronization module scheme.Finally,the serial sliding correlation operation is completed for the baseband sequence according to the frame synchronization code to realize frame synchronization.The above modules can all meet the system index requirements through simulation,which helps to complete the preprocessing scheme design before signal demodulation.Third,the system demodulation scheme of the AIS is designed and a differential Viterbi demodulation scheme is proposed.To start with,the traditional demodulation algorithm is studied,and the simulation and comparison of the differential phase detection algorithm are carried out,which aims at the burst communication mode of the AIS system.Additionally,the best demodulation scheme of GMSK signal for the coherent Viterbi demodulation algorithm based on maximum likelihood criterion is studied.The burst communication mode cannot meet the coherent demodulation requirements of the optimal demodulation scheme,and the residual frequency offset and carrier phase offset cause interference to the system,which in turn makes the coherent Viterbi demodulation algorithm collapse.Therefore,a two-bit differential Viterbi demodulation scheme is proposed in this thesis.Compared with the coherent Viterbi demodulation algorithm,this scheme expands the ability of anti-frequency offset and anti-phase offset,and reduces the requirements for front-end timing synchronization;Compared with the differential phase detection algorithm,it significantly improves demodulation and anti-frequency offset performance;Compared with the one-bit differential Viterbi demodulation algorithm,it reduces the introduction of noise and has better demodulation performance.This scheme adopts a new branch metric calculation method,which is calculated by the trigonometric function correlation value of the received signal.Compared with the absolute value difference of the differential phase angle,the demodulation performance of this scheme is better.In the end,the Viterbi demodulation algorithm is simulated and compared in different aspects to verify the feasibility of the scheme.Based on the above analysis,the scheme design of the AIS receiver signal processing system is completed.