Abstract: To estimate the angle signal in the high-dimensional signal processing efficiently, a second-order cone algorithm based on L₁ norm(L₁-SVD) is proposed. The algorithm applys sparse reconstruction to the target direction finding technology. Sparse domain model is introduced based on narrowband signal model,in which angle estimation problem of a high-dimensional signal is abstracted into underdetermined equations problems. Simulation results with Matlab show that, compared with other minimum norm methods and classic multiple signal classification algorithm, L₁-SVD can achieve lower reconstruction error and a higher probability of success in a wide range of signal to noise ratio and identify the signal with larger correlation. So it is proved that the algorithm can achieve the target source direction estimation effectively.

Key words: high-dimensional signal processing, L₁-SOC, sparse reconstruction, signal angle estimation