Wasserstein distance-based interference characterization and analysis in NGSO constellations with non-homogeneous stochastic geometry

Abstract

The performance degradation of mega Non-Geostationary Orbit (NGSO) satellite systems caused by co-frequency interference exhibits significant uncertainty, primarily due to the time-varying geometric topology, diverse communication signals, and dynamic channel conditions. Traditional approaches for characterizing NGSO interference often depend on non-statistical methods, which tend to incur substantial information loss. To address this issue, this paper proposes a co-frequency interference characterization model based on the Wasserstein distance and derives a closed-form expression for the ergodic capacity to support NGSO system performance analysis. Specifically, the NGSO constellation satellites are modeled as a Non-homogeneous Poisson Point Process (NPPP), while User Terminals (UTs) are modeled as a Homogeneous Poisson Point Process (HPPP). By incorporating a satellite-terrestrial Rician channel model and the orbital motion constraints of the NGSO constellation, an NGSO–NGSO co-frequency interference model is established. Furthermore, a chain mapping model is introduced to construct a topologically equivalent relationship among NGSO constellation configuration parameters, the statistical distribution of satellite signals, and the Wasserstein metric space. Finally, a closed-form expression for the system ergodic capacity is derived using higher-order stochastic geometry, revealing a negative correlation between the Wasserstein metric and the ergodic capacity. Simulation results validate the correctness of the chain mapping model and demonstrate the dual effectiveness of the Wasserstein distance in both interference characterization and system performance prediction.