Exploiting Timing Side-Channels in Quantum Circuits Simulation Via ML-Based Methods

2025-09-15

AI Summary

This work presents the first systematic demonstration of fine-grained timing side-channel vulnerabilities in cloud-based quantum circuit simulators: malicious co-located processes can infer the structure and identity of concurrent users’ private quantum circuits by monitoring execution-time patterns, posing a severe threat to quantum cloud privacy. To address this, we propose a machine learning–based joint timing–memory feature analysis method and conduct empirical evaluation using the QASMBench benchmark. Our approach achieves 88%–99.9% circuit identification accuracy across multiple datasets. The study not only empirically validates the practical feasibility of timing leakage in quantum simulators but also pioneers the application of side-channel analysis to the quantum software stack. By bridging classical security methodologies with quantum computing infrastructure, it establishes a novel analytical framework and provides concrete empirical foundations for secure design and mitigation strategies in quantum cloud platforms.

Technology Category

• Machine Learning: Quantum Machine Learning
• Computer Vision: Adversarial Attacks & Robustness
• Cognitive Modeling & Cognitive Systems: Simulating Human Behavior

Application Category

• Security and Privacy: Large-scale security measurements
• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Web Mining and Content Analysis: Web data generation and simulation

Abstract

As quantum computing advances, quantum circuit simulators serve as critical tools to bridge the current gap caused by limited quantum hardware availability. These simulators are typically deployed on cloud platforms, where users submit proprietary circuit designs for simulation. In this work, we demonstrate a novel timing side-channel attack targeting cloud- based quantum simulators. A co-located malicious process can observe fine-grained execution timing patterns to extract sensitive information about concurrently running quantum circuits. We systematically analyze simulator behavior using the QASMBench benchmark suite, profiling timing and memory characteristics across various circuit executions. Our experimental results show that timing profiles exhibit circuit-dependent patterns that can be effectively classified using pattern recognition techniques, enabling the adversary to infer circuit identities and compromise user confidentiality. We were able to achieve 88% to 99.9% identification rate of quantum circuits based on different datasets. This work highlights previously unexplored security risks in quantum simulation environments and calls for stronger isolation mechanisms to protect user workloads

Problem

• Timing side-channel attacks on cloud quantum simulators
• Extracting sensitive quantum circuit information via execution patterns
• Classifying circuit-dependent timing profiles to compromise confidentiality

Innovation

• Machine learning classifies quantum circuit timing patterns
• Timing side-channel attacks target cloud quantum simulators
• Co-located malicious processes extract sensitive circuit information