Abstract:

 A block sparse reconstruction algorithm based on the triple mixed norms is proposed to solve the problem of L-shaped nested array two-dimension angle estimation. Firstly, a two-dimension sparse direction finding model is established. In this model, the azimuth and elevation can be separated. The sampling points of the two dimensions are divided into blocks. The joint covariance matrixes are calculated separately. The two-dimension angle estimation problem is transformed into a joint covariance matrix sparse optimization problem. In order to reduce the computational complexity, a triple mixed norms block sparse reconstruction model is established. The sparse solution is achieved by the cross iteration method. The 2-D angle estimation is realized, and the angles of the two dimensions can be automatically matched. The simulation results show that the triple mixed norms sparse reconstruction algorithm can effectively estimate the two-dimension angle of the multiple sources, with high resolution and robustness. Compared with the traditional algorithm, at low SNR conditions and a small number of snapshots, the algorithm is superior to the traditional 2D angle estimation.

Key words: Nested array, Compressed sensing, Two-dimension angle estimation, Mixed norm, Block sparsity