Description
Date depot: 29 octobre 2023
Titre: Internet measurement techniques for measuring the impact of the centralization the Internet
Directeur de thèse:
Olivier FOURMAUX (LIP6)
Encadrant :
Timur FRIEDMAN (LIP6)
Encadrant :
Kevin VERMEULEN (Laas)
Domaine scientifique: Sciences et technologies de l'information et de la communication
Thématique CNRS : Systèmes et réseaux
Resumé: In an era of the centralization of the Internet by a few major content providers which deliver most of the traffic (sandvine report), it is important to think about the geopolitical implications of this evolution of the Internet structure. These hypergiants have built big private Wide Area Networks (WANs) to bypass the public Internet, gaining more control on how the traffic is delivered and improving performance. For the end user and external observer, this performance improvement came with a major drawback in terms of transparency: we have even less visibility than before on the paths taken by the traffic, making measurements more challenging. For instance, geolocating the routers on a path to infer the country-level path between two hosts of the Internet could be attempted via the combination of traceroute and multilateration measurements, but routers in a private WAN usually do not respond to traceroute packets. From these paths, one could answer questions related to the censorship performed by countries: recent work has designed techniques that leverage traceroute data to identify the equipment on the path that was performing censorship, leaving room for designing techniques to circumvent censorship in the presence of alternate paths. Another geopolitical concern with the centralization of the Internet is the impact on digital sovereignty. With the increasing dominance of cloud providers hosting and duplicating data from users all over the world, it is likely that sensitive data used by some websites and applications are replicated in multiple datacenters, leaving the opportunity for a foreign government to have a look at sensitive data belonging to users from other countries. In this thesis, we plan to design, implement, and deploy new measurement techniques to measure the impact of the centralization of the Internet. We do not restrict ourselves to the two applications of censorship and data sovereignty: Any other measurement technique that can reveal the geopolitical impact of the Internet is within the scope of this thesis. The value of such new measurement techniques and the datasets that they will produce go far beyond the scope of Internet measurements: They will help governments and policy makers by providing them with quantitative and qualitative data to better understand the geopolitical implications of the way in which the Internet is structured, and will also inform citizens about the privacy of their data on the Internet.
Doctorant.e: Rimlinger Hugo