This paper investigates the security strength of anonymous routing protocols in Network-on-Chip (NoC) architectures, a key component in multicore System-on-Chip (SoC) designs. The authors demonstrate that the existing anonymous routing is vulnerable to machine learning (ML) based flow correlation attacks. They propose a lightweight countermeasure that uses traffic obfuscation techniques to defend against these ML-based attacks with minor hardware and performance overhead. Experimental results show a high accuracy (up to 99%) for diverse traffic patterns, affirming the efficacy of their proposed attack and countermeasure.
Publication date: 28 Sep 2023
Project Page: Not provided
Paper: https://arxiv.org/pdf/2309.15687
