Benshan Mei

dblp:238/7713 · DBLP profile ↗
← Back
6ranked-venue papers
5as first author
3since 2021 · last 2025
0000-0002-2637-0101ORCID · corroborated

Domains — the database's venue-derived domains; a paper can count in several

Artificial intelligence and machine learning · 3 · 3 first-authorSecurity and privacy · 3 · 2 first-author · 3 since 2021
YearPublicationVenuePosition
2025 The Road to Trust: Building Enclaves within Confidential VMs
Wenhao Wang 0001, Linke Song, Benshan Mei, Shijun Zhao, Shoumeng Yan, XiaoFeng Wang 0001, Dan Meng 0002, Rui Hou 0001
NDSS3
2025 VMPL-KMI: Protecting Kernel Module Integrity within Confidential VMs
abstract
Confidential Virtual Machines (CVMs), such as AMD SEV, offer external protection but lack a privilege hierarchy, making them vulnerable to susceptible loadable kernel modules (LKMs). Although the Integrity Measurement Architecture (IMA) checks load-time integrity, it is prone to kernel exploits. While AMD SEV-SNP’s VM Privilege Levels (VMPLs) offer hardware-enforced intra-CVM isolation, they remain underexploited for kernel protection. This paper presents VMPL-KMI, utilizing VMPLs to enforce both load-time measurement and runtime protection of LKMs. By maintaining MList in the most privileged VMPL0 and synchronizing integrity checks with memory protection, VMPL-KMI prevents tampering and eliminates Time-Of-Check-to-Time-Of-Use (TOCTTOU) vulnerabilities. To minimize the overhead, we introduce a service protocol based on the Secure VM Service Module (SVSM) standard, reducing verification to a single interaction. Experimental results demonstrate practical efficiency: 5–10% overhead for small modules (over 20% for larger ones) during loading and less than 5% during unloading.
Benshan Mei, Wenhao Wang 0001, Dongdai Lin
TrustCom1
2024 Cabin: Confining Untrusted Programs Within Confidential VMs
Benshan Mei, Saisai Xia, Wenhao Wang 0001, Dongdai Lin
ICICS (1)1
2020 Safe sample screening for regularized multi-task learning
Benshan Mei, Yitian Xu
Knowl. Based Syst.1
2020 Multi-task ν-twin support vector machines
Benshan Mei, Yitian Xu
Neural Comput. Appl.1
2019 Multi-task least squares twin support vector machine for classification
Benshan Mei, Yitian Xu
Neurocomputing1