Nonlinear Beamforming in Digital Arrays for Weather Radar Networks

A weather radar network relies on two or more transmitters and receivers for localized weather event monitoring. For a receiving aperture implemented as an every-element digital array radar, it provides opportunity to leverage convex optimization for coordinated beamforming and null forming, in order to perform weather metrology without impairments. This work describes weather radar network challenges on receive due to ground clutter. The use of second-order cone programming, a scalable form of nonlinear optimization, mitigates clutter interference by deriving array element weights for beamforming in the target direction and null-forming in the ground clutter directions. Convex optimization enables beamforming with more flexibility when compared to conventional null steering with low sidelobe array tapers. This is demonstrated through quantifying the signal-to-clutter ratio impact of nulling. This paper demonstrates how digital beamforming via convex optimization in weather radar networks improves the signal-to-clutter ratio for a volume resolution cell at low altitude when compared to conventional beamforming methods. A figure of merit is presented to show how these clutter mitigations can improve wind field estimates for weather radar network applications such as 3D wind field retrieval via multi-Doppler methods in the presence of ground clutter. The paper concludes with an evaluation of a well-known adaptive digital beamformer for clutter mitigation as well as a comparison of all methods considered for a surveillance region of interest.