Dynamic Scheduling for the Satellite Range Scheduling Problem Using Greedy, MILP, Genetic Algorithm, and Deep Reinforcement Learning

Computational satellite scheduling is a critical challenge in modern space communication systems. This study investigates the Satellite Range Scheduling Problem (SRSP), which involves coordinating limited satellite-ground communication opportunities known as visibility windows. As satellite networks expand and communication demands become increasingly dynamic, static scheduling approaches become less effective. This research develops and evaluates a dynamic scheduling framework that integrates heuristic, optimization, and AI-based methods to handle new communication requests in real time. Using open orbital and ground-station data, the study quantifies measurable criteria such as throughput (MB), runtime (s), and schedule disruption (%) to compare algorithm performance under varying request frequencies. The goal is to establish a reproducible, quantitative framework for analyzing adaptive scheduling strategies in satellite communication systems.