Research On DOA Estimation Algorithm Of Coherent Distributed Source

Direction of Arrival(DOA)estimation is an important branch in the field of array signal processing.Classical estimation algorithms are all based on the assumption of point source.However,due to the complexity of the actual communication environment,scattering,reflection or diffraction are generated when the signal is transmitted under the influence of obstacles.As a result,the signal presents a certain spatial distribution and has a certain Angle expansion when reaching the antenna array.In application scenarios such as low elevation radar tracking and mobile communication,information sources are usually characterized as coherent distributed sources.Based on the coherent distributed source model,this paper studies Angle parameter estimation algorithm,and the specific work content is as follows:(1)The classical parameter estimation algorithms of coherent distributed sources are studied,including DSPE algorithm and TLS-ESPRIT algorithm.The basic principles of the two algorithms are discussed and simulated respectively.Simulation experiments show that the two algorithms can achieve effective angle estimation.DSPE algorithm has high estimation accuracy,while TLS-ESPRIT algorithm has the advantage of low computational complexity.(2)When the single target source is incident,the calculation complexity of the subspace algorithm is high because the distributed source has more estimation parameters of Angle expansion than the point source.In order to solve this problem,this paper firstly studies the fast algorithm based on propagation operator,and proposes a cascade OPM algorithm based on mutual quality array,which reduces the dimension of spatial spectrum search through cascade and makes the computational complexity significantly reduced.Aiming at this problem,this paper also adopts the idea of dimensionality reduction and proposes a dimensionality reduction DSPE algorithm.By separating the estimated parameters of the array guide vector,the cost function of the spatial spectral function is constructed,and the function relation between the central azimuth and Angle extension is established after its derivation.The estimation of the central azimuth and Angle extension is realized only by one-dimensional spatial spectral search,which can effectively simplify the computation.The estimation accuracy of this algorithm is the same as that of DSPE algorithm when the single target source is incident.(3)When multi-target information sources are incident,there are coherence problems between the sources.This paper studies the decoherence algorithms of dimensionality reduction and non-dimensionality reduction,and proposes a weighted space smoothing algorithm based on virtual array expansion.The conjugate virtual array increases the aperture of the array,uses an algorithm for pre-estimation,constructs a weighting matrix according to the pre-estimated value,the rank of the self-covariance matrix of the array response is restored by quadratic weighted spatial smoothing to achieve the purpose of coherence decoupling,and the incidence Angle of the target source is estimated by using the propagation factor algorithm.The algorithm expands the virtual array and makes full use of the auto-correlation information and cross-correlation information of the sub-array,thereby reducing the influence of spatial smoothing on the array aperture.The algorithm can accurately estimate the number of target sources,and has better estimation performance,and has better robustness at low signal-to-noise ratios.