Learning Emergent Random Access Protocol for LEO Satellite Networks

A mega-constellation of low-altitude earth orbit (LEO) satellites (SATs) are envisaged to provide a global coverage SAT network in beyond fifth-generation (5G) cellular systems. LEO networks exhibit extremely long link distances many users under time-varying topology. This makes existing multiple access protocols, such as random channel (RACH) based protocol designed for fixed terrestrial topology, ill-suited. To overcome this issue, paper, we propose novel contention-based solution networks, dubbed emergent (eRACH). In stark contrast model-based and standardized eRACH is model-free approach that emerges through interaction with the non-stationary environment, using multi-agent deep reinforcement learning (MADRL). Furthermore, by exploiting known orbiting patterns, does not require central coordination or additional communication across users, while training convergence stabilized regular patterns. Compared RACH, show from various simulations our proposed yields 54.6% higher average throughput around two times lower delay achieving 0.989 Jain’s fairness index.