Atomic norm method for DOA estimation in random sampling condition
Tong Qian, Jing Tian, Xu Zhang, Wei Li Cui · 2016
A novel direction-of-arrival (DOA) estimation method based on atomic norm in missing data condition is proposed. This method can exactly recover the missing data in random distribution and get the precise estimated result. A generalized atomic norm model with missing data is built to change the data recovery problem into a semidefinite programming issue. The dual problem and parameters selection are also analyzed to accelerate the proposed method. An enhanced covariance matrix sparse representation method is introduced for final DOA estimation. This method can deal with the coherent signal condition without any decorrelation preprocessing. Simulation experiments are conducted to validate the effectiveness of the proposed method and the results of several existing sparse representation based methods are also given for comparison.