Proposal on rain attenuation prediction method using convolutional neural network

抄録

Recently, the practical application of HAPS (High Altitude Platform Station) as the next-generation communication platform is studied actively. HAPS employs adaptive rain attenuation countermeasure techniques such as site diversity methods, therefore it is ideal to predict rain attenuation on the path in real time. We proposed real-time rain attenuation prediction method by convolutional neural network that inputs image of rainfall rate and path distance. Result showed that prediction accuracy of our proposed method is better than a method using conventional formulas.