Unified Security Framework Using Device-Specific Fingerprint: Mitigating Hardware Trojans and Authenticating Firmware Updates

System-on-Chips (SoCs) are ubiquitously used as smart devices with heterogeneous components embedded within them. The effective functioning of a system depends on Intellectual Property (IP) and firmware updates, which enhance performance and reliability and rectify vulnerabilities throughout the lifespan of any SoC. However, malicious logic intrusion by intruders can compromise the system’s efficacy at the design phase of Internet of Things (IoT) applications. The proposed robust unified security framework efficiently streamlines the identification of hardware Trojans and secure firmware updates of the IPs in all intricate and heterogeneous SoCs without the need for the golden reference model. In the post-silicon validation technique, the authentication protocol utilizes the device-specific fingerprint of a Physical Unclonable Function (PUF) to authenticate the IPs. An enhanced Balanced Anderson Arbiter PUF (BAA-PUF) integrating two sources of randomness is physically implemented in ZCU104 MPSoC to generate 256 bits. The parameters of BAA-PUF are determined by majority voting and are close to the ideal value. The average execution time for identifying the Trojan with the proposed framework is found to be 40.184 ms. Two metrics have been introduced to demonstrate the efficacy of the proposed framework, namely, utilization scale, which illustrates the impact of Trojans on the resource utilized and digest uniqueness, which demonstrates the presence of Trojans and ranges from 45.3% to 55%. The performance of the authentication protocol is also validated based on computation and communication costs. The authentication protocol is analyzed using the formal technique, employing Burrows–Abadi–Needham (BAN) logic and informal methods to confirm its robustness against hardware-based attacks.