CNN-based Weather Signal Detection Algorithm For Airborne Weather Radar

To detect weather signal submerged in ground clutter, a new approach is proposed for airborne weather radar by exploiting convolutional neural network (CNN) techniques in this paper. The information in Doppler domain and spatial domain is incorporated for the design of CNN and the detailed structure of the network is provided. The clutter phase alignment (CPA), Doppler velocity and interferometric phase in elevation are employed as the input of the CNN. Since we have not enough real data, especially labelled data, at hand, the proposed networks are now trained and tested via simulation radar echoes. As demonstrated by simulation results, the proposed algorithm overperforms most of the current weather signal detection methods under clutter background, and it can maintain good detection performance and good robustness under the condition that spectral moment information changes. Moreover, CNN show better classification performance than conventional classification networks such as Bayesian classifier and support vector machine.