A few projects in the Censored Planet Lab
The Censored Planet research lab at the University of Michigan investigates and exposes different types of privacy and security violations in the Internet. We design and deploy scalable techniques and systems to detect and protect users’ Internet experience from censorship, geo-discrimnation, third-party tracking, and surveillance.
Some of our projects include:
Censored Planet Censorship Observatory
Our observatory is a platform that provides continuous, global data about Internet censorship practices in countries around the world. The Censorship Observatory builds on our long line of work developing remote censorship measurement techniques. Our research group operates several of these systems, curates the data, and publishes semi-weekly datasets about the reachability of thousands of sensitive websites from more than 180 countries.
We plan to scale the project to a full-fledged production system that will have tools for monitoring data health, dynamically refresh vantage points to collect more comprehensive data and insights from next-generation remote censorship measurements available to the entire Internet Freedom community.
Reliably detecting interference
Reliably detecting interferences require significant effort in creating new frameworks for understanding and measuring network behavior. We design and develop techniques that bridge remote and in-situ measurement techniques, and use them to collect, analyze and distribute censorship measurement data of unprecedented scope and detail. Alongside, we develop empirical studies to understand network interference phenomenon. Examples of such techniques and studies are: HyperQuack, Quack, Satelite, Augur, Spookyscan etc.
Country specific studies
Sometimes to understand the nuances of security and privacy violations within specific countries, we perform interdisciplinary investigation into uncovering and exposing malpractices by different actors such as governments and service providers. An example of such study is how the Russian government is gradually building national-level censorship policies on thousands of privately owned ISPs using inexpensive commodity middle-boxes, a trend that we fear other countries with similar topological structure will follow. Examples of such studies are: Russia, Kazakhstan MiTM, Great Cannon, Great Firewall of China etc.
Investigating dual-use technologies
Security threats have driven rapid progress in the sophistication and deployment of network security technologies, such as firewalls, deep-packet inspection middleboxes (DPIs), and intrusion detection systems (IDSs). Studying how these technologies are configured and applied, and where they are deployed, can help us better understand how to monitor and defend against interference. An example of such a study is our FilterMap, a novel framework for discovering and monitoring DPI-based filtering at Internet scale, based on the blockpages DPI systems display.
The Internet is becoming increasingly regionalized due to sanctions, financial regulations, copyright and licensing rights, perceived abuse, or a perceived lack of customers. For instance, the majority of the news has been either on geoblocking of multimedia products or geolocation-based price discrimination. We lack a global perspective on the extent of this phenomenon and we are looking at the multitude of ways that netizens can be discriminated while using the Internet based on their location. An example , we performed the first wide-scale measurement study of geoblocking; our data showed that certain practices by content-delivery networks (CDN) likely contributed to overly aggressive blocking by their customers, resulting in entire national populations being unable to reach valuable sites and content.
Web privacy and transparency
We look into how web protocols are deployed and are evolving, while collecting empirical evidence on how websites and 3rd party services collect user data.