The Research Of Technology For Super-Resolution Direction Finding To Coherent Signals

Technology for counteracting attack of Anti Radiation Missile(ARM) has been developped rapidly in recent years, one of them is setting interference system consisting of active decoy and radar. Probability of being detected of radar system is reduced by using decoy, in order to improve capability of counteracting decoy and angle resolution of the passive detector, theory of super-resolution Direction-of-arrival(DOA) estimation for coherent signal is studied in this dissertation, including source number estimation, fast implementation of dircection finding based on virtual array, utility of polarization sensitive array, optimization of estimation method for wide-band signal, each subject is studied in detail, and corresponding calibration methods are proposed, and some indexes are tested in the computer, the results prove the proposed methods valid.The excellent performance of spectral estimation algorithm is under the assumption of knowing the source number precisely, so the correct estimation of source number is very important to achieve precise DOA estimation. Source number estimation methods under different conditions are studied. Coherent source number can be estimated by smoothing-rank sequence method under the Gaussian white noise background, it is convenient and effective; An approach is presented to determine number of coherent signals under the colored noise background, it combined information theoretic criteria and eigenvalue correction, the method uses Alternating Projection Method and information theoretic criteria to estimate angles and number of coherent signals alternately, then eliminate the inequality of the eigenvalues caused by colored noise by correcting the noise eigenvalues, consequently, even number of coherent signals can be estimated in colored noise preferably, simulation results illustrate the effectiveness of the new method.Uniform Circle Array(UCA) is a array in common use, the antennas distributing on a circle uniformly, it is superiority to uniform linear array. But its array manifold is complex, the expresstion is not Vandermonde form like uniform linear array, so some excellent algorithms based on Uniform Linear Array(ULA) can not be used on UCA directly. The performance of the virtual uniform linear array is analyzed and Mode-Space Smoothing Algorithm is researched based on transform of Mode-Space. Finally, a new algorithm called Mode-Space ESPRIT is proposed. It makes decorrelation between signals by Mode-Space reconstruction, then estimates the result by ESPRIT, the method need not spatial smoothing and spectrum searching, consequently, the calculation is reduced effectively, the simulation results show the effectiveness of the method.Only one field information of the signal can be acquired by routine scalar sensor array, the information used is delay of the phases between sensors, it can not acquire the polarization information of signal. Electromagnetic vector sensor array offsets the deficiency of the scalar sensor array, most of the decorrelation algorithms based on scalar sensor array only make use of the airspace information, the polarization information and characteristics of cyclo-stationaray signals is not made the best. Whereas most manpower signals are cyclo-stationary, the temporal speciality of signal is used here, a new algorithm called cyclo-stationary coherent signal based on polarization sensor array is proposed. The experiment results show that the method can restrain disturbance and noise background effecttively, and it has the signal selectivity by cyclo-frequency, and decorrelat the signal by the polarization information effectively.Wide-band signal is researched, model of the wide-band signal is introduced, then CSM(Coherent Signal-Subspace Method) is analyzed, then one of the algorithm—TCT(Two-sided Correlation Transformation) method is discussed, finally fast approach to 2-D wide-band coherent signal is proposed. The signal subspace and noise subspace are acquired by directly decomposing the covariance matrix of the received data, then getting the focus matrix by signal subspace of every frequency, and greatly reduces the amount of computation and the consumed timed to favor the real-time application of wide-band DOA estimation, meanwhile the performance of TCT method is preserved.