Zoom Meeting: https://umass-amherst.zoom.us/j/9621986782
The Internet has emerged as one of the most important tools of communication. With around 4.5 billion active users as of July 2020, it provides people the opportunity to access a vast treasure trove of information and express their opinions online. However, some countries consider the Internet as a critical communication medium and attempt to deploy network interference strategies. National governments, in particular, are notorious for their attempts to impose restrictions on online communication. Further, certain Internet service providers (ISPs) have been known to throttle specific applications and violate net neutrality principles.
Alongside the proliferation of network interference, and an increasing awareness of the security and privacy of users over the Internet, we have seen a rise in the usage of network traffic encryption technologies. However, even with encryption enabled, network interference is still possible due to the information leakage of the DNS and TLS protocols. To this end, a rich ecosystem of DNS/TLS improvements has come to light with the purpose of improving user privacy by obfuscating the domains a user visits. These protocols have the potential to render certain forms of censorship ineffective.
In this thesis proposal, I will describe my contributions towards understanding network interference including, Internet censorship, as well as throttling of specific network applications (traffic differentiation). I develop a network measurement platform that enables monitoring of network interference globally on an ongoing basis. I also investigate the prevalence of traffic differentiation practices and how they impact popular video streaming applications. Additionally, I present my work on understanding the DNS censorship behavior of the Great Firewall of China (GFW). Then I will review two recent proposals, namely DNS over HTTPS/TLS (DoH/DoT), and Encrypted Server Name Indication (ESNI) that aim to improve the security and privacy of the DNS and TLS protocols and investigate their potential to mitigate network interference and improve user privacy.
For the remaining portion of my thesis, I plan to develop network measurement techniques to study the prevalence of domain name encryption technologies and their impact on Internet Censorship longitudinally, around the world. Additionally, I am also continuing my research on the GFW to better understand its design in DNS censorship by monitoring it over time and to detect poisoned responses. This can aid in sanitizing poisoned DNS caches and assist in developing circumvention tools to bypass the GFW's DNS censorship.
Advisor: Phillipa Gill