A Churn for the Better: Localizing Censorship using Network-level Path Churn and Network Tomography
This addresses the issue of unintended international censorship leakage for global Internet users, providing a systematic study at a global scale.
The paper tackled the problem of identifying which autonomous systems (ASes) perform censorship on the Internet, using BGP churn and network tomography to narrow down potential censoring ASes by over 95% and exactly identifying 65 censoring ASes, with 24 of these causing censorship leakage affecting users outside their jurisdiction.
Recent years have seen the Internet become a key vehicle for citizens around the globe to express political opinions and organize protests. This fact has not gone unnoticed, with countries around the world repurposing network management tools (e.g., URL filtering products) and protocols (e.g., BGP, DNS) for censorship. However, repurposing these products can have unintended international impact, which we refer to as "censorship leakage". While there have been anecdotal reports of censorship leakage, there has yet to be a systematic study of censorship leakage at a global scale. In this paper, we combine a global censorship measurement platform (ICLab) with a general-purpose technique -- boolean network tomography -- to identify which AS on a network path is performing censorship. At a high-level, our approach exploits BGP churn to narrow down the set of potential censoring ASes by over 95%. We exactly identify 65 censoring ASes and find that the anomalies introduced by 24 of the 65 censoring ASes have an impact on users located in regions outside the jurisdiction of the censoring AS, resulting in the leaking of regional censorship policies.