Mahdi Rahimi (KU Leuven)

Mix networks (mixnets) provide clients with communication anonymity against strong network adversaries by traversing their packets independently through randomly selected hops (mixnodes), which disrupt packet linkability. Although this approach, implemented in Nym, maximizes obfuscation against network adversaries, it enables an adversary who compromises a subset of mixnodes ($10$%/$5$% of nodes) to entirely nullify the anonymity of all clients whose communication volume with their destination exceeds a certain threshold ($4$MB/$30$MB).

To mitigate such vulnerabilities, this work develops a set of novel path selection techniques that achieve a trade-off between resistance to network adversaries and resilience against compromised mixnodes. Observing that existing anonymity metrics are insufficient to quantify adversarial risk in mixnets, we additionally introduce effective empirical and simulation-based metrics.
Through theoretical, empirical, and simulation-based evaluations, we comprehensively assess our proposals, demonstrating that the proposed approaches reduce the vulnerability to compromised nodes by up to $80%$, while conferring limited advantage to network adversaries. Our analysis further reveals that state-of-the-art anonymity metrics, in contrast to our proposed metrics, produce misleading results that influenced certain design choices in Nym.

View More Papers

Hey there! You are using WhatsApp: Enumerating Three Billion...

Gabriel Karl Gegenhuber (University of Vienna), Philipp Frenzel (SBA Research), Maximilian Günther (University of Vienna), Johanna Ullrich (University of Vienna), Aljosha Judmayer (University of Vienna)

Read More

SECV: Securing Connected Vehicles with Hardware Trust Anchors

Martin Kayondo (Seoul National University), Junseung You (Seoul National University), Eunmin Kim (Seoul National University), Jiwon Seo (Dankook University), Yunheung Paek (Seoul National University)

Read More

Benchmarking and Understanding Safety Risks in AI Character Platforms

Yiluo Wei (The Hong Kong University of Science and Technology (Guangzhou)), Peixian Zhang (The Hong Kong University of Science and Technology (Guangzhou)), Gareth Tyson (The Hong Kong University of Science and Technology (Guangzhou))

Read More