AT-Simulator: An Aviation Traffic Simulation Framework for LEO Communication Networks

With the rapid development of Low Earth Orbit (LEO) satellite constellation communication networks, providing internet access services to aviation users has emerged as a key application. However, the complex and dynamic nature of network demands in aviation scenarios poses challenges for traditional traffic prediction models, which rely heavily on historical data. To address these limitations, we introduce AT-Simulator, a novel aviation traffic simulation framework designed for LEO networks. AT-Simulator leverages real-time flight trajectory data and H3 Cell technology to generate precise aviation traffic distribution data, minimizing the dependence on historical information. The framework features an innovative preprocessing method that models flight trajectories through geometric sampling, along with a dynamic gateway planning and traffic matching mechanism. By simulating aircraft positions over time and using a geodesic-based flight-to-gateway matching algorithm, an aviation traffic (AT)-Simulator effectively predicts network demands across various periods. Experimental validation using real flight data and several ground stations demonstrates the framework's capability to accurately capture the spatial-temporal characteristics of aviation traffic, providing valuable insights for network measurement and traffic forecasting in LEO satellite constellations.