SC-GAN: A spectrum cartography with satellite Internet based on Pix2Pix generative adversarial network

The increasing demand for radio authorized applications in the 6G era necessitates enhanced monitoring and management of radio resources, particularly for precise control over the electromagnetic environment. The radio map serves as a crucial tool for describing signal strength distribution within the current electromagnetic environment. However, most existing algorithms rely on sparse measurements of radio strength, disregarding the impact of building information. In this paper, we propose a spectrum cartography (SC) algorithm that eliminates the need for relying on sparse ground-based radio strength measurements by utilizing a satellite network to collect data on buildings and transmitters. Our algorithm leverages Pix2Pix Generative Adversarial Network (GAN) to construct accurate radio maps using transmitter information within real geographical environments. Finally, simulation results demonstrate that our algorithm exhibits superior accuracy compared to previously proposed methods.