Monte Carlo Throughput Estimation in Dynamic LEO Satellite Networks

Abstract

This study introduces a comprehensive framework for analyzing capacity dynamics and throughput performance in low Earth orbit satellite networks (LSNs) with unreliable intersatellite links (ISLs). It addresses two critical limitations of prior models: (1) the lack of capacity and throughput evaluation under dynamic network conditions, an inherent and unavoidable characteristic of LSNs, and (2) the inability of flow-network models (e.g., maximum flow) to estimate throughput under specific, practical routing schemes, as they inherently assume traffic is routed according to an optimal flow distribution. To address these challenges, we propose the capacity model under a dynamic LSN (Cap-DLSN), which characterizes the time-varying link capacity under stochastic ISL failures, along with a Monte Carlo throughput estimation (MCTE) framework that evaluates network throughput under dynamic traffic and diverse routing policies. Simulations on operational constellations show that MCTE provides realistic and computationally efficient performance benchmarks, offering practical guidance for routing optimization or network design in next-generation LSNs.