Orbital Prediction Based Routing for ISL-Enabled LEO Satellite Networks

Low Earth Orbit (LEO) satellite networks, such as Starlink, are undergoing remarkable development. However, related research has reported that communication in these networks suffers from link quality fluctuations and performance degradation due to frequent handovers. These frequent handovers stem from the limited communication range in conventional bent-pipe environments. Consequently, Inter-Satellite Links (ISLs) are being explored to expand this range. However, a practical route selection method for ISL-enabled environments has not yet been established, and protocols like AODV result in massive control overhead. To address this issue, this paper proposes a new route selection method. We extend a technique from our previous work, which predicted the trajectories of individual satellites, to a new approach that first predicts the trajectory of individual satellites on the same orbit and then selects the optimal satellites for the ISLs route. This method is designed to provide long communication durations in ISL-enabled environments. We validated its effectiveness on the ns-3 LEO module to simulate a future satellite constellation. The results demonstrate that, compared to AODV, proposed method significantly increases communication duration by 222.6% and dramatically reduces control overhead, although this comes at the cost of reduced route diversity.