Research On Direction Of Arrival Estimation Algorithms For Airspace Signals Using Sparse Signal Reconstruction

Direction of arrival estimation is one of the most important research directions in the field of array signal processing.It has a wide range of applications in radar detection,wireless communication and satellite navigation.In today's world,the demand for technology is increasing.The traditional direction-of-arrival estimation algorithm can't meet people's requirements in terms of accuracy,snapshot or anti-interference ability.With the continuous development of the theory of compressed sensing in these years,the method of using signal sparsity to perform sparse reconstruction has been successfully applied in the direction of arrival estimation,which promotes the further development of the direction-of-arrival estimation technology.This paper will study the direction of arrival estimation of airspace signals based on sparse reconstruction techniques.For the far-field narrow-band signals in the space domain,this paper focuses on two kinds of spatial domain signal estimation algorithms based on sparse reconstruction techniques,which are sparse representation-based DOA estimation algorithm and sparse iteration-based DOA estimation algorithm.1.For the DOA estimation algorithm based on sparse representation,this paper introduces?_1SVD and?_1SRACV algorithms,and analyzes the role of the meshing of the airspace in the sparse representation process of the covariance matrix of the observed signals.The idea of converting the?₀ norm problem into the?₁ norm problem is proposed.For the limitation that the?_1SVD algorithm needs to predict the number of signal sources and the computational complexity of the?_1SRACV algorithm is too large,we propose the KR-SRACV algorithm,which transforms the covariance matrix of the observed signal into the Khatri-Rao product.The form of the vector is used for sparse reconstruction to simplify the computational complexity of the algorithm.2.For the DOA estimation algorithm based on sparse iteration,this paper introduces a sparse iterative DOA estimation algorithm based on covariance matrix,and analyzes it to construct a new augmented steering matrix.According to the covariance fitting criterion,it establish an optimization equation to solve the sparse spectrum of the target signal power matrix.Similar to the solution in the?_1SVD algorithm,the Frobenius norm problem is transformed into a semi-definite programming problem to solve in the process of solving the optimization equation.Then we introduce the elastic net model and propose an improved EN-SPICE algorithm.According to the characteristic of elastic net model that can select the variables,this model can be applied to the constraint term of the optimization equation of the power matrix sparse spectrum in the SPICE algorithm.The correlation between the noise interferences is weakened,which can be exploited to reduce the impact of noise interference in the covariance matrix of the observation matrix,so that the sparse spectrum obtained by the solution will be more accurate.