Planar antenna arrays of a balanced structure are of great importance in many applications due to their features including low-profile, wideband and high polarisation purity. However, feeding such antennas adds great complexity in terms of manufacturing and reduces the performance due to production of common mode propagation. A method for feeding such antennas using coplanar waveguide on a thin substrate is proposed. Not only does it terminate the antenna of a balanced structure with a single-ended feed but it also enables impedance transformation. The design was an attempt towards a completely printed front-end that incorporates the antenna elements and their feeding circuits on bendable substrates. The electromagnetic performance has been validated with a dual polarized prototype and its prospect for ultrawideband arrays e.g., 5G sub-6 GHz or square kilometre array, was explored.

Standard convolution is difficult to provide an effective fog feature for visibility estimate tasks due to the fixed grid kernel structure. In this paper, a multiscale deformable convolution model (MDCM) is proposed to extract features that make effectively sampling discriminating features from the atmospheric region in foggy image. Moreover, to enhance performance we use RGB-IR image pair as observations and design a multimodal visibility range classification network based on the MDCM. Experimental results show that both the robustness and accuracy of visibility estimate performance are raised beyond 30% compared to standard convolutional neural networks (CNNs).