Investigating Web Content Delivery Performance over Starlink

2025-10-15

AI Summary

This study investigates the impact of Starlink’s low-Earth-orbit (LEO) satellite internet on Web content delivery performance, challenging the conventional CDN paradigm predicated on geographic proximity and stable terrestrial network paths. Leveraging multi-source active measurements—including Cloudflare AIM, M-Lab, RIPE Atlas, and custom probes—we analyze 6.1 million traceroutes and 10.8 million DNS queries across network, transport, and application layers, augmented by causal inference. We identify three infrastructure-density-driven performance patterns, demonstrating that physical proximity of content nodes—not satellite coverage per se—is the dominant factor governing latency. Our findings expose critical mismatches in DNS resolution and CDN mapping logic under LEO networks, necessitating architectural rethinking. Empirically, following Starlink’s 2025 capacity expansion and strategic content node decentralization, median page load time decreased by 60%, and RTT in previously high-latency regions dropped below 200 ms.

Technology Category

• Search and Optimization: Distributed Search
• Application Domains: Mobility, Driving & Flight
• Data Mining & Knowledge Management: Web

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Web performance, measurement, and characterization
• Web Mining and Content Analysis: Web measurements
• Responsible Web: Measurement, analysis, and circumvention of Web censorship

Abstract

Low Earth Orbit (LEO) satellite ISPs promise universal Internet connectivity, yet their interaction with content delivery remains poorly understood. We present the first comprehensive measurement study decomposing Starlink's web content delivery performance decomposed across Point of Presence (PoP), DNS, and CDN layers. Through two years of measurements combining 225K Cloudflare AIM tests, M-Lab data, and active probing from 99 RIPE Atlas and controlled Starlink probes, we collect 6.1M traceroutes and 10.8M DNS queries to quantify how satellite architecture disrupts terrestrial CDN assumptions. We identify three distinct performance regimes based on infrastructure density. Regions with local content-rich PoPs achieve near-terrestrial latencies with the satellite segment dominating 80-90% of RTT. Infrastructure-sparse regions suffer cascading penalties: remote PoPs force distant resolver selection, which triggers CDN mis-localization, pushing latencies beyond 200 ms. Dense-infrastructure regions show minimal sensitivity to PoP changes. Leveraging Starlink's infrastructure expansion in early 2025 as a natural experiment, we demonstrate that relocating PoPs closer to user location reduces median page-fetch times by 60%. Our findings reveal that infrastructure proximity, not satellite coverage, influences web performance, requiring fundamental changes to CDN mapping and DNS resolution for satellite ISPs.

Problem

• Analyzing web content delivery performance over Starlink satellite networks
• Quantifying how satellite architecture disrupts terrestrial CDN assumptions
• Identifying infrastructure proximity as key factor for web performance

Innovation

• Decomposed Starlink web performance across PoP DNS CDN layers
• Identified three performance regimes based on infrastructure density
• Demonstrated PoP relocation reduces page-fetch times by 60%