Research On Optimal Covariance Broadband DOA Estimation Based On Modified Matrix Decomposition

At present,in the requirement of international communication system technology,besides large capacity and high rate,how to improve the signal bandwidth and spectrum efficiency is also one of the difficult problems.Compared with narrow-band signal,broadband signal carries more information,and its resolution and anti-jamming ability is stronger than narrow-band signal,and the delay spread and multipath attenuation of broadband signal can be effectively overcome by the use of array technology.Therefore,compared with narrow-band signal,broadband signal has many advantages,which makes the research of broadband signal of great value.Therefore,in this paper,the direction of arrival for wideband array signals is studied.The subspace class algorithm of rival,DOA is studied,and the algorithm is improved according to the influence of signal-to-noise ratio,beat number and non-uniform noise.The main research work is as follows:Firstly,the array signal is modeled and the parameters are analyzed.In this paper,the multiple classification algorithm for narrow band signal DOA estimation is studied.The principle and existing problems of MUSIC algorithm and spatial smoothing algorithm which can realize coherent signal de-coherence are studied and analyzed.The subspace algorithm of wideband array signal is studied.The wideband signal subspace class DOA algorithm is divided into incoherent signal subspace?ISM?algorithm and coherent signal subspace?CSM?algorithm.This paper introduces the principle of ISM algorithm and its existing problems.However,the research of CSM class algorithm mainly focuses on how to construct the focus moment.In this paper,three methods of constructing focus matrix are introduced and compared with each other.To solve the problems that the traditional direction-of-arrival?DOA?estimations have poor performance when processing coherent signals in the cases of low signal-to-noise ratio and non-uniform noise.Based on the modified matrix decomposition method,with the best covariance matrix by the convex optimization.A DOA method is proposed.The modified matrix decomposition method can deal with the extraction of the coherent sources,while,the aperture loss caused by the traditional algorithm is overcomed.Furthermore,the noise free covariance matrix is reconstructed by the convex optimization.Finally the DOA estimation can be calculated by the minimized search.The simulation results show that,comparing to the matrix decomposition?MD?algorithm,l₁-norm singular vector decomposition?l₁-SVD?algorithm and spatial smoothing based covariance rank minimization?SS-CRM?algorithm,the method suppress the non-uniform noise well and have good performance in low signal noise ratio?SNR?.Finally,the TOPS algorithm can overcome the dependence on the accuracy of angle pre-estimation for the problem that CSM algorithm can achieve de-coherence but is too dependent on the pre-estimation of incident angle,and the performance of TOPS algorithm can deteriorate under the condition of non-uniform noise.However,there are still too many pseudo-peaks in the algorithm,which can not solve the interference of non-uniform noise on the performance of DOA estimation.Based on the further improvement of TOPS algorithm,an improved TOPS algorithm based on OC-SMD algorithm is proposed.In order to solve the problem of excessive pseudo peaks in TOPS algorithm,the construction matrix of the algorithm is modified.Then,based on the convex optimization principle,the covariance is used to solve the problem.