Research On The Separation Algorithm Of AIS Signal And ADS-B Signal

Automatic Identification System and Automatic Dependent Surveillance-Broadcast system provide real-time navigation and monitoring management services for maritime traffic and air traffic respectively.Self-organized time division multiple access technology can avoid conflicts between AIS devices in the same cell.However,time slot conflicts are likely to occur when signals arrive at the receiver in spaceborne AIS.ADS-B system transfers information between the airplane unit and the base station by random access broadcast message.The signal is sent independently at all times,which leads to aliasing.Firstly,this article introduces the frame structure as well as coding and modulation method of AIS signal and ADS-B signal.Then it illustrates the study of three different types of separation algorithms for the aliased signals with the same frequency in the ADS-B system and in the AIS.The main work is included as the following:(1)In order to solve the problem of aliased signals with the same frequency,it is proposed to use array signal processing to separate the signals.Firstly,multiple signal classification or rotation invariance technology is used to estimate the direction of arrival of the desired signal and the interference signal.Then,weight vector of the array aligns is adjusted according to beamforming,which ensures that the main lobe of the array is aimed at the target signal.At the same time,interference and noise is suppressed.The simulation results show that the array signal processing is able to recover the target signal although outstanding performance depends on accurate angle estimation and array flow pattern matrix.(2)Aiming at the signal model where the AIS signals and the ADS-B signals do not completely overlap,a blind adaptive beamforming algorithm is proposed.The blind beamforming algorithm gets rid of the dependence on the prior knowledge of the direction of the source signals and the array flow pattern,which accounts for good robustness.With the help of alternating direction multiplier method,the optimal vector is sought by solving the nonconvex optimization problem.For the same frequency mixed signals without complete overlap,the separation results have good signal-to-noise ratio performance.(3)Independent Component Analysis algorithm plays a vital role in the solution of blind source separation problem.Considering the second-order non-circularity of AIS signal and ADS-B signal,the non-circular complex Fast ICA algorithm is derived based on the complex Fast ICA algorithm;at the same time,considering the mixed model affected by noise,the observation data is projected into the signal subspace,which effectively removes the influence of noise.The simulation results show that for noise-containing mixed AIS signals and mixed ADS-B signals,the non-circular Fast ICA algorithm and denoising non-circular Fast ICA can separate both circular and non-circular signals without any prior knowledge.