Super-resolution Direction Finding at Subarray Level for Coherent Sources Based on Weighting Network

2-D subarray level super-resolution direction finding methods have important applications in phased array radars. This paper studies ML (Maximum Likelihood) method at subarray level suitable for coherent sources and gives the corresponding signal model. Applying simplified array manifolds can make calibration cost of phased array reduced largely. We post-process the digital subarray outputs by bringing in weighting network which increases the flexibility of array processing greatly. The simplified array manifold based on Gaussian patterns we constructed can overcome the limitations of DSAM (Direct Simplified Array Manifold) method that available direction estimation area can't be changed and the uninterested sidelobe sources can't be suppressed completely, but the cost is that the precision of direction finding drops. Simulation results demonstrate the validity of the proposed method.