Prof. Yongsoo Joo - Kookmin University AI·SW Graduate School

Research Overview 연구 개요

Our research focuses on building high-performance, energy-efficient, and secure storage systems. We develop I/O completion optimization techniques that dynamically adapt to workload characteristics, achieving both prompt and CPU-efficient storage access on modern SSDs and NVMe devices. Our work spans from firmware-level security (ransomware-proof SSDs) to OS-level optimization (virtualized I/O, application launch acceleration), with results published at top-tier venues including USENIX FAST and ACM HotStorage.

본 연구실은 고성능, 에너지 효율적, 안전한 스토리지 시스템 구축을 목표로 연구하고 있습니다. 워크로드 특성에 동적으로 적응하는 I/O 완료 최적화 기술을 개발하여, 최신 SSD 및 NVMe 장치에서 신속하고 CPU 효율적인 스토리지 접근을 달성합니다. 펌웨어 수준의 보안(랜섬웨어 방지 SSD)부터 OS 수준의 최적화(가상화 I/O, 앱 실행 가속)까지 폭넓은 연구를 수행하며, USENIX FAST, ACM HotStorage 등 최상위 학회에 논문을 게재하고 있습니다.

Research Areas 연구 분야

I/O Completion Optimization
Storage Security
Virtualized I/O Performance
SSD/NVMe Performance Optimization
Energy-Efficient Memory Systems

I/O 완료 최적화
스토리지 보안 (랜섬웨어 방어)
가상화 I/O 성능 최적화
SSD/NVMe 성능 최적화
에너지 효율적 메모리 시스템

Major Achievements 주요 연구 성과

Dynamic I/O completion method (DPAS) achieving 9% YCSB improvement on 3D XPoint SSD
Polled I/O path expansion for Linux virtualization
Memory-efficient ransomware detection at SSD firmware level
28% reduction of application launch time on SSDs (FAST)
Hybrid PCM/Flash storage architecture for performance-energy-lifetime co-optimization

동적 I/O 완료 기법(DPAS)으로 3D XPoint SSD에서 YCSB 9% 성능 향상 달성
리눅스 가상화 환경의 Polled I/O 경로 확장
SSD 펌웨어 수준의 메모리 효율적 랜섬웨어 탐지
SSD 기반 애플리케이션 실행 시간 28% 단축 (USENIX FAST)
PCM/Flash 하이브리드 스토리지 아키텍처로 성능-에너지-수명 동시 최적화

Honors & Awards 수상

US Patent 11,768,532 — I/O Latency Tracking for Storage Devices, 2023
KR Patent 10-2919645 — Hybrid Interrupt Processing, 2026
US Patent App. 2026/0030187 — I/O Completion Processing, 2026

Recent Publications 주요 논문

D. Seo, J. Jung, Y. Yoon, P.-X. Chen, Y. Joo, S.-S. Lim, and N. Dutt, "DPAS: A Prompt, Accurate and Safe I/O Completion Method for SSDs," USENIX FAST, 2026
D. Seo, Y. Joo, and N. Dutt, "Improving Virtualized I/O Performance by Expanding the Polled I/O Path of Linux," ACM HotStorage, pp. 31-37, 2024
Y. Joo, Y. Yoon, and J. H. Choi, "A Memory-Efficient Overwrite Detection Method for Ransomware-Proof SSDs," IEICE Trans. on Info. and Systems, Vol. 106, Issue 8, 2023
Y. Joo, D. Seo, D. Shin, and S.-S. Lim, "Enlarging I/O Size for Faster Loading of Mobile Applications," IEEE Embedded Systems Letters, Vol. 12, Issue 2, 2020
Y. Joo, S. Park, and H. Bahn, "Exploiting I/O Reordering and I/O Interleaving to Improve Application Launch Performance," ACM Trans. on Storage, Vol. 13, Issue 1, 2017
Y. Joo, J. Ryu, S. Park, and K. G. Shin, "FAST: Quick Application Launch on Solid-State Drives," USENIX FAST, pp. 259-272, 2011
G. Sun, Y. Joo, Y. Chen, Y. Xie, Y. Chen, and H. Li, "A Hybrid Solid-State Storage Architecture for the Performance, Energy Consumption, and Lifetime Improvement," HPCA, pp. 1-12, 2010
Y. Joo, J. Ryu, S. Park, H. Shin, and K. G. Shin, "Rapid Prototyping and Evaluation of Intelligence Functions of Active Storage Devices," IEEE Trans. on Computers, Vol. 63, Issue 9, 2014