Securing RISC-V by Reinforcing Execution with Instruction-Level Control Flow Integrity for Attack-Resistant Computing

This work introduces a Securing RISC-V(SecRISC-V) by Reinforcing Execution with Instruction-Level Control Flow Integrity (IL-CFI) for Attack-Resistant Computing, SecRISC-V (IL-CFI) is a method to secure execution using Instruction-Level Control Flow Integrity to enhance computing systems' attack resistance. The research entails: control flow verification of instructions at instruction level to ensure adherence to defined execution paths, anomaly detection using a Cosine Similarity-Centric Convolutional Neural Network (CS-C2Net) to identify abnormal patterns, optimization of the network parameters using Addax Optimization (AdxO) to obtain optimal detection accuracy and implementation using Fibonacci Q-Matrix Hyperchaotic Encryption (FiQ-MHE) to secure control flow information. Experimental results confirm the efficacy of the method, providing control flow verification accuracy of 99.5%, anomaly detection accuracy of 99.7%, encryption integrity guarantee of 99.9%, and runtime overhead reduction by over 99.5%. Overall, SecRISC-V(IL-CFI) is a solid and effective method of securing RISC-V architectures, thus encouraging the development of more attack-resistant computing systems.