Small Language Model-based Control for BBR over Low Earth Orbit Satellite Internet

2026-07-08

AI Summary

This work addresses the challenge that Bottleneck Bandwidth and Round-trip propagation time (BBR) congestion control exhibits high retransmission rates and unstable throughput in low Earth orbit (LEO) satellite networks due to its aggressive bandwidth probing, struggling to balance high throughput with robustness. To overcome this limitation, the paper introduces, for the first time, small language models (e.g., SmolLM2, GPT-2) into satellite congestion control, proposing an adaptive framework driven by real-world Starlink measurements. The approach leverages structured BBR state encoding, efficient LoRA-based fine-tuning, and a constrained output head to learn safe rate adaptation policies directly from empirical link traces. Experimental results demonstrate that the proposed method preserves BBR’s high-throughput advantage while significantly reducing retransmission overhead, achieving performance comparable to large-model solutions at substantially lower computational cost.

Abstract

Low Earth Orbit (LEO) satellite Internet introduces rapid path variability, intermittent capacity shifts, and non-terrestrial delay dynamics that challenge transport-layer congestion control. Although Bottleneck Bandwidth and Round-trip propagation time (BBR) achieves high throughput in such environments, its aggressive bandwidth probing can cause excessive retransmissions and unstable pacing over LEO links. This paper presents a global experimental evaluation of BBR over a SpaceX Starlink testbed spanning six geographically distributed AWS endpoints and compares its behaviour against Cubic, Vegas, and Hybla under isolated and competing traffic scenarios. The measurements show that BBR consistently delivers superior throughput but incurs significantly higher retransmission overhead, revealing a critical throughput-stability trade-off in LEO satellite Internet. To address this limitation, we propose a Small Language Model (SLM)-guided BBR adaptation framework that learns phase-safe pacing-gain decisions from real Starlink traces. The framework combines a structured BBR state encoder, LoRA-based parameter-efficient fine-tuning, and a constrained networking head to generate feasible pacing actions with low inference latency. Evaluation using GPT-2, T5, GPT-Neo, and SmolLM2 shows that lightweight SLMs can retain BBR's throughput advantage while substantially reducing retransmissions, with performance comparable to larger language models but at much lower computational cost.

Problem

• LEO satellite Internet
• BBR
• congestion control
• throughput-stability trade-off
• retransmissions

Innovation

• Small Language Model
• BBR congestion control
• LEO satellite Internet
• pacing-gain adaptation
• parameter-efficient fine-tuning