Potential-Minimized Path First: A Novel Load-Balancing Routing Scheme for Mega Constellation Networks

For mega-constellation networks (MCNs), balancing network congestion and signaling overhead under high overloading is a critical challenge. This letter presents potential minimized path first (PMPF), a distributed potential-field-based routing policy where congestion notifications are selectively sent depending on their impact on network performance. The potential field effectively estimates the shortest path delay using ephemeris information and mimics the delay increase due to congestion with a concise potential model. With each satellite optimizing its routing decision based on potential minimization, signaling overhead is reduced by prioritizing the congestion notifications that significantly impact the potential. Also, routing stability is improved by updating the routing table only when it would significantly improve network performance. Simulation results show PMPF's potential to outperform a global state routing algorithm by up to 43% with less than 8% signaling messages, demonstrating its superior performance in joint scheduling among links.