| The Automatic Identification System(AIS)is considered to be the cornerstone of maritime traffic safety.It can realize the exchange of information between ships to ensure the safety of maritime traffic,the safety of people’s property and the balance of the aviation ecosystem.Spaceborne AIS serves as a hub for long-distance ship communications.It is of great significance to the safe navigation of ships at sea.According to the actual departure process of satellite-borne AIS mixed signal separation and detection,the problem of blind separation and detection of random burst and same-frequency mixed signals under single channel is studied.The main work of this paper is as follows:(1)The classical Viterbi algorithm is studied,which provides important theoretical support for guiding the blind separation and detection process of signals.The essence of Viterbi algorithm is the optimal solution problem of multi-step and multi-selection model.The selection at each step saves the possibility of all previous steps to the current step.After all the steps are calculated in turn,the optimal selection path is found by the backtracking method.Per-Survivor Processing(PSP)embeds data-aided unknown parameter estimates into the Viterbi algorithm to achieve joint estimation of parameters and sequences.This can be applied to the symbol sequence of the original signal from the received signal under conditions where the parameters are unknown.The process parameters are more advantageous,so this principle can also be used.(2)A joint estimation algorithm for single-channel AIS mixed signal separation and detection based on improved PSP is proposed.The algorithm breaks the traditional PSP execution step and accumulates the surviving symbol state into the symbol state set of the detection process during the blind separation symbol transfer process.This combines the blind separation and detection process.The joint estimation algorithm breaks the traditional signal processing method,and the improved method can directly output the received mixed signal as the information sequence of the original signal.(3)A joint estimation method based on PSP for AIS mixed signal separation and detection is proposed.The method is based on the characteristic that the relative delay of the AIS is large,and the amplitude signal detection method is used to divide the mixed signal into an aliased segment and an unmixed segment.For the aliasing segment,PSP blind separation is performed first,and then the separated sequence is subjected to PSP detection.For the non-aliased segment,PSP detection is performed directly.The original signal is finally stitched back according to the specific frame format of the AIS signal.The simulation results show that the proposed method can combine the two processes of separation and detection with stable performance,high estimation accuracy and low complexity.(4)An aliasing position estimation algorithm for mixed signals based on double window detection is proposed.The algorithm proposes an energy detection method to estimate the aliasing position for the problem of large error and high complexity of the aliasing position estimation of mixed signals.A double sliding window detection method for detecting the burst position of a burst signal is improved.The original dual energy detection window is augmented into four energy detection windows and the decision function is improved.This method only uses the signal energy to estimate the aliasing position,which reduces the influence of other parameters on the algorithm.The simulation results show that the improved algorithm performance is obviously better than the frequency amplitude detection method,and the robustness is strong and the complexity is low.The magnitude of the estimation result is close to the Cramerome boundary. |
