A paper titled “Enhancing Network Attack Detection with Distributed and In-Network Data Collection System” is accepted at USENIX Security 2024. This paper explores a new possibility of distributed feature measurement framework in the era of AI using data plane programmability. The idea is to prioritize large flows over super mice to measure immediately in the local switch while passing super mice over to the next hop switch. This approach resolves the resource scarcity issues to enhance both the flow coverage and the measurement integrity.

Congratulations to Mehdi, Ashwin, David, and especially RhongHo!