Research On Direction Of Arrival Estimation Algorithm Based On Coprime Array

Direction of arrival(DOA)technology is commonly used in radar search,sonar detection and other fields,and is widely used in wireless communication engineering scenes.For DOA estimation of incoherent signals,representative algorithms include beamforming algorithms and subspace algorithms.The angular resolution of beamforming algorithm is low,and the subspace method is limited by Nyquist sampling theorem,and the estimation accuracy is poor for low signal-to-noise ratio environment.For the DOA estimation of coherent signals,the traditional estimation methods for independent signals fail,and the coherent signals need to be dephased first.The commonly used preprocessing methods include spatial smoothing algorithm and reconstruction algorithm.However,this preprocessing is always accompanied by the loss of array aperture.In addition,the fitting algorithm can complete the coherent signal estimation without losing the array aperture,but the setting of the initial value of this kind of algorithm has a great impact on the estimation accuracy of the algorithm,and the anti-interference performance is poor.Therefore,a coprime array that can expand the array aperture is introduced.Compared with uniform linear array with same array element,the DOA estimation algorithm based on coprime array has lower array installation cost and better performance in low signal-to-noise ratio environment.After analyzing the classical DOA estimation methods,two DOA estimation methods for incoherent and coherent signals based on coprime array are proposed,and compared by simulation experiments.It is verified that coprime array has the unique advantages of expanding array aperture.The main work includes:1.The similarity between traditional array signal model and coprime array signal model is analyzed.Based on this similarity,coprime sampling can break through the limitation of Nyquist sampling theorem.The principle of coprime sampling is extended to virtual domain of signal processing,and the excellent performance of virtual domain extension of coprime array is verified.2.When the incident signal is incoherent,the beamforming algorithm,subspace algorithm,coprime array deblurring algorithm and virtual domain expansion algorithm based on binary coprime array are analyzed.It is proved that ternary coprime array has a larger virtual domain than binary coprime array with the same number of elements.In order to maximize the use of the virtual domain of Coprime array,an algorithm based on the expansion of the virtual domain of ternary coprime array is proposed.Firstly,the weighted average of the covariance elements with the same wave path difference in the covariance matrix of ternary coprime array is reconstructed into Toeplitz matrix,and then the largest continuous array of Toeplitz matrix in the virtual domain is divided into overlapping virtual subarrays,and the covariance matrices of each subarray are weighted and averaged to enhance the signal,Then the subspace algorithm is used to estimate the DOA of the reconstructed smooth covariance matrix.The simulation results show that compared with the traditional algorithm,the proposed algorithm can not only complete more DOA estimation with the same number of array elements,moreover,it has high estimation accuracy in the environment of low signal-to-noise ratio.3.When the incident signal is coherent,the preprocessing algorithms of spatial smoothing,fitting and reconstruction are simulated and analyzed.Aiming at the problems of serious array aperture loss and poor estimation performance in low signal-to-noise ratio environment,a DOA estimation method based on forward Toeplitz reconstruction of Coprime array virtual array is proposed.Firstly,the zero element of the antenna is filled into the virtual array of the coprime array to expand the array aperture;Then,the covariance matrix is reconstructed into Toeplitz matrix to enhance the signal robustness;Then,the reconstructed array is smoothed forward and backward to eliminate the signal coherence;Finally,the cost function is designed to transform the angle parameter solution into a convex optimization problem,and the DOA estimation is obtained by spectral peak search.The simulation results show that compared with the traditional decoherence algorithm,the proposed algorithm does not lose the array aperture,has strong anti-interference ability,and does not need the prior information of the number of sources.