BISSEL: Built-In Self Security via Embedded Sensors for Reproducible Side-Channel Leakage Assessment

Abstract

Adversaries continue to advance Side-Channel Analysis (SCA) techniques, making these threats increasingly sophisticated and impactful. Such attacks are now recognized as official CVEs (Common Vulnerabilities and Exposures), highlighting their disruptive potential. A key challenge remains: how can the vulnerability of an implementation to side-channel attacks—stealthy and dependent on how they are carried out—be effectively assessed? Two main groups grapple with this question: evaluators, who require reproducible modus operandi methods for security assessments, and designers, who need to validate the robustness of their implementations. While normative frameworks like ISO/IEC 20085 provide guidance, their systematic application in practice remains unclear. This paper addresses the need for valid and measurable quantification of SCA vulnerabilities in chips. We propose leveraging integrated sensors as an alternative to noisy external side-channel observations. Embedded sensors can serve both designers and evaluators, enhancing Systemon-Chip (SoC) security while requiring safeguards to prevent their misuse as attack vectors. We demonstrate how aggregating data from multiple embedded sensors enables low-cost leakage assessments, supporting a built-in self-security evaluation method (BISSEL). The BISSEL approach has been rigorously validated through FPGA implementation and extensive experimentation, showcasing its effectiveness and robustness.