29 Jan 2024 | Ferhat Erata, Chuanqi Xu, Ruzica Piskac and Jakub Szefer
This paper presents the first formalization and demonstration of using power traces to unlock and steal quantum circuit secrets. The work shows how attackers can recover information about the control pulses sent to quantum computers, which can then be used to reverse-engineer the gate-level description of circuits and eventually the secret algorithms. Two new types of single trace attacks are introduced: per-channel and total power attacks. The per-channel attack uses per-channel measurements to perform a brute-force attack to reconstruct the quantum circuits, while the total power attack uses Mixed-Integer Linear Programming optimization to perform a single-trace attack. The work demonstrates that quantum circuit secrets can be stolen with high accuracy. Evaluation on 32 real benchmark quantum circuits shows that the technique is highly effective at reconstructing quantum circuits. The findings highlight the need to develop new means to protect quantum circuits from power trace attacks. The paper also discusses the significance of vulnerabilities in quantum computer controllers, drawing lessons from historical technological threats. It emphasizes the importance of safeguarding quantum computing systems and the need for proactive security measures. The paper provides a formalization of power side-channel attacks on quantum circuits, demonstrating how attackers can reconstruct quantum circuits from power traces. The work introduces a novel algebraic reconstruction method for recovering quantum circuits and evaluates the attacks on 32 real quantum circuits in the QASMBench benchmark suite. The paper also discusses the background of quantum computers, including qubits, quantum gates, control pulses, and pulse-level circuit descriptions. It outlines the attack scenario and threat model, assuming attackers can sample power traces from shots of a circuit or measure a single total power trace. The paper also discusses the attacker's objective, which is to uncover quantum circuit details from captured power traces. The impact of attacks is discussed, highlighting the importance of protecting intellectual property in quantum computing. The paper concludes by emphasizing the need for proactive security measures and the importance of formal methods in enhancing quantum computing security.This paper presents the first formalization and demonstration of using power traces to unlock and steal quantum circuit secrets. The work shows how attackers can recover information about the control pulses sent to quantum computers, which can then be used to reverse-engineer the gate-level description of circuits and eventually the secret algorithms. Two new types of single trace attacks are introduced: per-channel and total power attacks. The per-channel attack uses per-channel measurements to perform a brute-force attack to reconstruct the quantum circuits, while the total power attack uses Mixed-Integer Linear Programming optimization to perform a single-trace attack. The work demonstrates that quantum circuit secrets can be stolen with high accuracy. Evaluation on 32 real benchmark quantum circuits shows that the technique is highly effective at reconstructing quantum circuits. The findings highlight the need to develop new means to protect quantum circuits from power trace attacks. The paper also discusses the significance of vulnerabilities in quantum computer controllers, drawing lessons from historical technological threats. It emphasizes the importance of safeguarding quantum computing systems and the need for proactive security measures. The paper provides a formalization of power side-channel attacks on quantum circuits, demonstrating how attackers can reconstruct quantum circuits from power traces. The work introduces a novel algebraic reconstruction method for recovering quantum circuits and evaluates the attacks on 32 real quantum circuits in the QASMBench benchmark suite. The paper also discusses the background of quantum computers, including qubits, quantum gates, control pulses, and pulse-level circuit descriptions. It outlines the attack scenario and threat model, assuming attackers can sample power traces from shots of a circuit or measure a single total power trace. The paper also discusses the attacker's objective, which is to uncover quantum circuit details from captured power traces. The impact of attacks is discussed, highlighting the importance of protecting intellectual property in quantum computing. The paper concludes by emphasizing the need for proactive security measures and the importance of formal methods in enhancing quantum computing security.