|Title||Min Dong-hyun and Ahn Jin-woo, Graduate School Students Led by Prof. Kim Young-jae in Computer Science and Engineering, Selected for Best of IEEE CAL Award with their Thesis|
Min Dong-hyun and Ahn Jin-woo, Graduate School Students
Led by Prof. Kim Young-jae in Computer Science and Engineering,
Selected for Best of IEEE CAL Award with their Thesis
The thesis jointly written by Min Dong-hyun and Ahn Jin-woo, graduate students in Computer Science and Engineering, was selected for the 2018 Best of IEEE CAL Award. The IEEE Computer Architecture Letters (CAL) is a world-renowned journal with a thesis selection rate of approximately 24%.
The result of the studies was announced at the Best of CAL session in the 25th International Symposium on High-Performance Computer Architecture (HPCA 2019), the world’s best academic seminar in the fields of computer architecture, held in Washington D.C. on February 16-20.
The title of the paper selected as the Best of IEEE CAL Award is “Amoeba: An Autonomous Backup and Recovery SSD for Ransomware Attack Defense,” (Primary Author: Min Dong-hyun, Responsible Author: Prof. Kim Young-jae).”
The study is a proposal on solid-state drive (SSD) technology that enables automatic data backup and restoration against ransomware attacks.
The core technology of the study is the firmware S/W technology in SSD for automatic backup restoration against ransomware attacks and the H/W DMA technology for detecting high-speed ransomware. The SSD developed from this study can independently detect data fabrication attacks without any support from the host computer, and the attacked data is automatically restored in backup so that the source data can be restored when attacked by ransomware. By applying machine learning technology, the ransomware detection technology proposed by the study can accurately detect ransomware attacks with a very high percentage.
The study was conducted jointly by the next generation system S/W lab at the Computer Science and Engineering Department, led by Prof. Kim Young-jae, the decentralized cloud lab, led by Prof. Park Sung-yong, and Prof. Lee Jung-hee’s lab at the Electrical and Computer Engineering Department at the University of Texas San Antonio (UTSA).
At the next generation system S/W lab at Sogang, the SSD firmware algorithm software for automatic backup and restoration was developed, and the decentralized cloud lab developed a kernel module to extract the I/O trace including block I/O content in the Linux kernel block layer. On the other hand, the UTSA research team designed the hardware DMA engine for high-speed ransomware detection.