LeoTCP: Low-Latency and High-Throughput Data Transport for LEO Satellite Networks

2025-08-26

AI Summary

Low Earth Orbit (LEO) satellite networks face three critical challenges: non-congestive latency spikes induced by satellite mobility and link fluctuations, transient hotspot congestion, and frequent connection disruptions due to handovers. To address these, this paper proposes a novel transport protocol leveraging in-network telemetry. Departing from conventional end-to-end signaling, it innovatively employs per-hop real-time telemetry—such as link quality, queue occupancy, and forwarding delay—to enable fine-grained, low-overhead dynamic rate control and coordinated path scheduling. Evaluated via OMNeT++/INET simulations and micro-benchmarks under path handover, non-congestive packet loss, and hotspot isolation scenarios, the protocol demonstrates significant improvements over state-of-the-art approaches: 42% higher effective throughput, 37% lower end-to-end latency, and 51% reduced buffer occupancy. These results confirm substantial gains in transmission robustness and resource efficiency under highly dynamic LEO network conditions.

Technology Category

• Planning, Routing, and Scheduling: Plan Execution and Monitoring
• Application Domains: Mobility, Driving & Flight
• Search and Optimization: Learning to Search

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Web performance, measurement, and characterization
• Security and Privacy: Large-scale security measurements
• Responsible Web: Data and user privacy-enhancing technologies for the Web

Abstract

Low-Earth Orbit (LEO) satellite networks consist of thousands of satellites orbiting the Earth, enabling low-latency and high-throughput communications across the globe. Such networks present unprecedented challenges due to their dynamic nature, which state-of-the-art data transport protocols do not address. These challenges include: (1) non-congestive latency variation and loss, caused by continuous satellite movement and fluctuating link quality due to weather effects; (2) transient hotspots leading to buffer build-up, latency inflation, and potential packet loss; and (3) frequent handovers, which may result in temporary connectivity loss and re-routing through paths with unknown congestion and delay characteristics. In this paper, we introduce LeoTCP, a novel data transport protocol designed specifically to address these challenges. LeoTCP leverages in-network telemetry (INT) to gather congestion information on a per-hop basis. Using this information, LeoTCP (1) minimises both buffer occupancy and latency for end users, (2) maximises application throughput and network utilisation, and (3) swiftly reacts to network hotspots. We compare LeoTCP to state-of-the-art data transport protocols using a LEO satellite simulation model and targeted micro-benchmarks, both based on OMNeT++/INET. The simulation model captures RTT dynamics in a simulated LEO satellite constellation, while the micro-benchmarks isolate key LEO-specific characteristics, including non-congestive latency variation and loss, path changes, and congestion hotspots. Our results demonstrate that LeoTCP significantly increases goodput compared to existing state-of-the-art approaches, while simultaneously minimising latency.

Problem

• Addressing non-congestive latency variation and loss in LEO networks
• Mitigating transient hotspots causing buffer build-up and latency
• Handling frequent handovers with unknown congestion characteristics

Innovation

• Uses in-network telemetry for congestion monitoring
• Minimizes latency and buffer occupancy dynamically
• Swiftly reacts to network hotspots and changes