Research Of Separation Algorithms For Space-Based AIS Signals

Automatic Identification System(AIS)is a kind of wireless communication system.It is used to communicate between ship to ship and ship to shore.Space-based AIS receives information through the satellite,at the same time,feedbacks the information to the test center of the shore.Then the long distance communication is realized.The received signals mix in time and frequency domain,close on space.So received signal separation is a problem worth studying.In this paper,according to the actual requirements of Space-based AIS signals,the separation problem under different conditions is studied.The main contents include:(1)A method of Space-based AIS signals separation is proposed.The method based on the characteristic of the doppler-frequency shift makes the received signal into multiple narrow band signals by filter,so that we can improve the rate of information detection.The simulation results show that the performance of four band filter is best and it will loss to a minimum.(2)FastICA algorithm is sensitive to the initial value and the robust performance is poor.Aimed at the problem,an improved FastICA algorithm is proposed.Considering the AIS signal constant model,the improved algorithm improves the initial vector through the stochastic gradient constant modulus criterion,and chooses Modified-M estimation function as objective function algorithm of nonlinear function so as to improve the robustness of algorithm.The improved algorithm also modifies Newton iterative algorithm,makes it meet the third order convergence,thus reduces the number of iterations and improves the convergence speed.The experimental simulation results show that improved algorithm is better on the separation accuracy and efficiency.(3)A new method of AIS signal extraction is proposed.The AIS mixed signal is decomposed into different Intrinsic Mode Function(IMF)by Empirical Mode Decomposition(EMD).Then the AIS mixed signal and the IMF form the new multidimensional signal.After that we can extract signal by the FastICA.At last,considering the characteristic of AIS signal,we use the Hilbert transform to process the extracted signal.Experimental simulation results show that the method can extract the high power signal.And if the frequency offset is bigger,the extracted signal is closer to the source signal.