Securing hardware coprocessors against piracy using biometrics for secured IoT systems
Aditya Anshul, Rahul Kumar Chaurasia, Anirban Sengupta · Institution of Engineering and Technology eBooks · 2023
Two crucial aspects need serious consideration concerning the underlying hardware used on the Internet of Things (IoT) systems: security and design optimization. Pirated or counterfeited hardware integrated into IoT systems may lead to different security threats such as confidential data leakage, malfunctioning of the device, and performance degradation. Therefore, the security of underlying hardware in IoT systems must be ensured for safe and reliable usage to end-users. This chapter discusses biometric-based methodologies to secure IoT hardware coprocessors against piracy and isolate counterfeited designs. Biometric-based approaches, such as fingerprint, facial, and palmprint biometrics, are discussed regarding their design flow, security strength, design cost overhead, tamper tolerance ability, and probability of coincidence (Pc). The biometric security in corresponding IoT hardware coprocessors is incorporated during the high-level synthesis phase (during the higher abstraction level of design) of the design process to ensure minimal design overhead and easy isolation of counterfeited designs. Moreover, for an adversary, the exact regeneration of digital templates to evade counterfeit detection is not possible due to incorporating unique biometric features during signature generation, unlike non-biometric-based approaches. The palmprint biometric hardware security approach renders higher tamper tolerance than the facial and fingerprint approaches (due to the larger encoding bits and signature strength), whereas the fingerprint biometric approach offers a smaller Pc than the facial and palmprint hardware security approaches.