Poster Abstract: Time Attacks using Kernel Vulnerabilities

📅 2025-03-27
📈 Citations: 0
Influential: 0
📄 PDF
🤖 AI Summary
This paper identifies system time maintenance—a long-overlooked attack surface—and demonstrates that kernel time subsystem vulnerabilities enable highly stealthy, privilege-escalated time manipulation attacks. Method: The authors formally define the “time attack” paradigm, systematically discover kernel vulnerabilities, and reverse-engineer clocksource/clockevent mechanisms to construct privileged time-tampering primitives, achieving millisecond-precise, controllable time offsets on mainstream Linux systems. Contribution/Results: Such attacks bypass critical security checks—including NTP synchronization validation, TLS certificate validity enforcement, short-lived token authentication, and chronological log auditing—exposing fundamental flaws in time-dependent security designs. The work establishes the first benchmark for time security assessment and introduces a comprehensive defense framework, thereby elevating time integrity to a first-class dimension of system security.

Technology Category

Application Category

📝 Abstract
Timekeeping is a fundamental component of modern computing; however, the security of system time remains an overlooked attack surface, leaving critical systems vulnerable to manipulation.
Problem

Research questions and friction points this paper is trying to address.

Investigates timekeeping security vulnerabilities in modern computing
Explores kernel-level time attack surfaces and risks
Addresses system time manipulation threats to critical systems
Innovation

Methods, ideas, or system contributions that make the work stand out.

Exploiting kernel vulnerabilities for time attacks
Manipulating system time as attack surface
Targeting overlooked timekeeping security flaws
🔎 Similar Papers
No similar papers found.
M
Muhammad Abdullah Soomro
University of Massachusetts, Amherst
Adeel Nasrullah
Adeel Nasrullah
University of Massachusetts, Amherst
Cyber Physical SystemsNetworked Embedded SystemsApplied Machine Learning
F
Fatima Muhammad Anwar
University of Massachusetts, Amherst