국내기사
RISC-V IoBT 환경의 자원고갈 공격분석 및 AI 기반 하드웨어 방어 메커니즘 연구 = Analysis of resource exhaustion attacks and AI-driven hardware defense mechanism in RISC-V IoBT environments
This dissertation proposes a lightweight AI-driven Hardware Intrusion Detection System (H-IDS) for SWaP-constrained RISC-V IoBT environments. Gem5 SE simulations established that resource exhaustion attacks induce structural bottlenecks, increasing CPI by approximately 7.4%. Leveraging HPM data, the Random Forest model achieved statistically significant detection performance by identifying these subtle architectural patterns. By demonstrating autonomous hardware-level defense capabilities without software overhead, this study provides a core foundation for securing national defense systems.