The Research On DOA Estimation Algorithm Under Different Array Models

Direction of Arrival(DOA)estimation is one of the research contents of array signal processing.It has a wide range of applications in military and economic fields such as radar,sonar,communications,and earthquakes.Nested arrays and coprime arrays are new types of arrays that have emerged in recent years.They can significantly improve the freedom of the array,handle DOA estimation when the number of sources is greater than the number of array elements,and improve estimation performance.In view of the limitations of the classical DOA estimation algorithm under different array models,the main research contents of this article are as follows:First of all,under the uniform linear array model,in order to solve the problem of low efficiency of the MUSIC algorithm when the large linear array or the number of samples is large,a Root-MUSIC algorithm based on Nystr(?)m and spectral factorization algorithm is proposed.This algorithm uses the Nystr(?)m method to decompose the signal receiving matrix to construct an approximate noise subspace.This process avoids the process of directly calculating the covariance matrix and the covariance matrix feature decomposition,and optimizes the MUSIC algorithm.The spectral factorization algorithm further reduces the polynomial order of Nystr(?)m-Root-MUSIC and reduces the computational complexity of Nystr(?)m-Root-MUSIC.Simulation experiments show that this algorithm not only has better estimation accuracy under higher signal-to-noise ratio,but also has higher estimation efficiency under large array environment.Then,under the nested array model,in order to solve the problem of poor performance of the classic DOA estimation algorithm when the number of elements in the virtual array is much larger than the original number of elements,a DOA estimation algorithm based on spectral factorization under the nested array is proposed.Based on the analysis of the nested array structure,this algorithm uses the Propagator Method algorithm to block the signal covariance matrix of the virtual array to construct an approximate noise subspace.The spectral factorization algorithm is further used to simplify the root-finding process of the PM-Root-MUSIC polynomial,and the performance of the PM-Root-MUSIC algorithm under the nested array is improved.Simulation results prove that this algorithm has low complexity and can perform highprecision DOA estimation.Finally,under the coprime array model,in order to improve the accuracy and realtime performance of the two-dimensional MUSIC algorithm under the environment of low signal-to-noise ratio and small snapshot,an improved two-dimensional MUSIC algorithm based on subspace projection is proposed.This algorithm uses the characteristics of non-circular signals to expand the signal receiving matrix,effectively converts the two-dimensional estimation problem into a one-dimensional estimation problem,and then further uses the subspace projection algorithm to modify the noise subspace.The simulation results show that this algorithm is in low confidence Better estimation performance in the environment of noise ratio and small snapshots.