RISC V Based Fault Diagnosis Architecture for Skinny Family of Block Ciphers

Cybersecurity vulnerabilities continue to exist in the world of cryptographic hardware and software, with risks originating from the environment and malevolent actors attempting to obtain access to secret keys. Lightweight cryptographic algorithms are typically used by resource-constrained devices to strike a compromise between security and effectiveness. Striking a balance between resource optimization and robust security measures remains a continual challenge. Fault injection attacks are one type of attack vector that has become much more sophisticated in recent years. For lightweight cryptography, choosing the best and most effective fault diagnosis systems is therefore essential.This work employs Error Detection via the Recomputation with Inverted Operands approach to improve the reliability of the SKINNY Block Ciphers using the RISC V architecture. The results show that the proposed fault detection method keeps SKINNY's throughput overhead within a 2.5% variance, with a maximum area overhead of 10% when employing a pipelined strategy. Our implementation of the fault detection scheme showcases a unique contribution, as no error detection-based fault detection technique has been previously proposed for the SKINNY block ciphers in the existing works for RISC V architecture. The implementation results validate the efficiency of the proposed system, particularly suited for resource-constrained environments.