Work

VIMA: A Privacy-Preserving Integrity Measurement Architecture for Containerized Environments

Year: 2025

Type: article

Abstract: Integrity verification and attestation are critical in containerized environments, where traditional Linux Integrity Measurement Architecture (IMA) falls short due to its lack of container-specific contextualization. These gaps undermine container autonomy, escalate privacy risks, and impede granular integrity checks. Addressing these challenges, this paper introduces the Virtual IMA (VIMA), a novel framework that refines Linux IMA’s principles to support containerized settings. Using nested Merkle trees, VIMA’s Two-Tree Architecture (2TA) enables detailed integrity assessments across system-wide monolithic trees and individual container trees. Integrating Merkle and zero-knowledge (ZK) proofs establishes VIMA as a secure, privacy-preserving verification and attestation solution. Our comparative analysis and initial prototype testing reveal that VIMA significantly improves upon traditional IMA with minimal performance overhead, offering substantial scope for optimization.

Source: IACR Transactions on Cryptographic Hardware and Embedded Systems

DOI: 10.46586/tches.v2025.i4.1053-1076

Authors: Omar Jarkas, Ryan K L Ko, Naipeng Dong, Redowan Mahmud

Institutions: The University of Queensland, Curtin University

Language: English

Topic: Security and Verification in Computing

Subfield: Artificial Intelligence

Field: Computer Science

Domain: Physical Sciences

Open Access status: diamond

OpenAlex ID: W4414017052

FWCI: 0.408

Cites:

Cited by: 1

Related to: 10