Angular resolution limit of two closely-spaced point sources based on hypothesis testing

In this paper, we consider the statistical angular resolution limit (ARL) on resolving two closely-spaced point sources in array processing based on the framework of hypothesis testing. A more general formulation of the linearized hypothesis test is proposed and a new analytical expression of the ARL is derived. The result is more general and its superiority over previous work is verified via numerical simulations. For the first time, we also investigate the case of two identical sources where a second-order approximation is necessary. Numerical simulations show that the theoretical result agrees with the Cramér-Rao bound (CRB) criterion-based ARL with regard to the relation between the resolution limit and the signal-to-noise-ratio (SNR).