Systematic Security Analyses of QUIC

• Applicant: Professor Dr.-Ing. Juraj Somorovsky
• Subject Area: Security and Dependability, Operating-, Communication- and Distributed Systems
• Term: since 2025
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 556138177

Cryptographic protocols play a crucial role in securing the Internet. They are used, for example, to protect connections to websites, emails, and establishing secure messaging applications. To evaluate and strengthen cryptographic protocol implementations, research in the area of applied cryptographic attacks has been established. One prominent target of such analyses is Transport Layer Security (TLS). Given the protocol's importance, TLS received much attention in the last decade. Security researchers evaluated TLS with theoretical analyses, searched for novel cryptographic attacks manually, and used state learning and large-scale scans to detect new vulnerabilities automatically. These efforts have led to the discovery of novel vulnerabilities and to the overall security enhancement of the TLS ecosystem. QUIC is a new cryptographic protocol designed with the motivation to speed up the connections and ease the deployment process. It is the transport layer for the new HTTP/3 protocol. QUIC is already in an advanced deployment state, which is mainly driven by large providers. According to the Cloudflare statistics website, more than 30% of Cloudflare connections are served using QUIC. Despite its importance, systematic security analyses of QUIC are still lacking. Cryptographically, QUIC is based on TLS 1.3; it utilizes the TLS 1.3 handshake and incorporates several TLS 1.3 concepts. One might assume that recent TLS analyses and their results directly apply to QUIC. However, this is not the case. QUIC introduces new concepts, differentiating it from TLS. Most notably, QUIC operates on the unreliable UDP layer, requiring the developers to handle connection unreliabilities, which are handled through a two-layer system of packets and frames. Consequently, many new QUIC libraries have been implemented from scratch, which introduces the potential for new cryptographic vulnerabilities. Given QUIC's importance and the lack of systematic security analyses, our goal is to develop novel security analysis techniques. Our project will rely on manual cryptographic analyses and interdisciplinary methods, especially state learning and large-scale Internet scanning. Our techniques aim to detect new vulnerabilities in local environments and on a large scale, enhancing the overall security of the QUIC ecosystem.

• DFG Programme: Research Grants