LyndHub
Jadwal Sholat

Memuat jadwal sholat…

← All journalsSearch index

Ilmu Komputer & AI editorial

Open AccessOA2026

Experimental Side Channel Analysis of Protocol Stages in Quantum Identity Authentication

Machine learning classification of protocol stages from optical side channels in a quantum communication testbed
Marwan Elawady; Lance Young; Contessa Wilburn; Blaine Keyton; Carrie Houston; Mohamed Shaban; Muhammad Ismail· 2026· DOI 10.48550/arXiv.2607.24639

The core problem

Quantum networks promise secure distributed computing and sensing, but their security relies on quantum identity authentication to prevent man-in-the-middle (MitM) attacks. Without authentication at the quantum layer, malicious repeaters can retain entanglement instead of performing swapping, enabling MitM attacks between communicating parties. Authentication mitigates this by embedding authentication qubits within data qubits at positions and bases determined by a secret key shared a priori. While prior work analyzes security and MitM detection guarantees, physical layer side channel analysis remains unexplored. If an attacker infers protocol stages, it can avoid authentication qubits and extract data qubits, rendering authentication ineffective. This paper addresses this gap by experimentally investigating whether protocol stages can be inferred from side channel observations in a quantum communication testbed.

Innovation

The experimental results demonstrate that protocol-stage inference is feasible with high accuracy. At 30% sampling, the classification accuracy reached 98% with an -score of 97%. At 10% sampling, the accuracy was 96% with an -score of 94%. These results indicate that even with a relatively small portion of the signal diverted, an attacker can reliably determine the protocol stage. The high performance across both sampling settings suggests that side channel information is rich enough to distinguish between different stages of the quantum identity authentication protocol. The findings are summarized in the following table:

| Sampling Rate | Accuracy | -Score |
|---------------|----------|----------|
| 30% | 98% | 97% |
| 10% | 96% | 94% |

These results highlight the vulnerability of the protocol to side channel inference attacks.

Quantum networks promise secure distributed computing and sensing, but their security relies on quantum identity authentication to prevent man-in-the-middle (MitM) attacks. Without authentication at the quantum layer, malicious repeaters can retain entanglement instead of performing swapping, enabling MitM attacks between communicating parties. Authentication mitigates this by embedding authentication qubits within data qubits at positions and bases determined by a secret key shared a priori. While prior work analyzes security and MitM detection guarantees, physical layer side channel analysis remains unexplored. If an attacker infers protocol stages, it can avoid authentication qubits and extract data qubits, rendering authentication ineffective. This paper addresses this gap by experimentally investigating whether protocol stages can be inferred from side channel observations in a quantum communication testbed.
The authors conducted experimental studies using a quantum communication testbed. A beam splitter was used to tap a portion of the optical signal, allowing an observer to collect side channel data without disrupting the quantum state. Two sampling settings were evaluated: 30% and 10% of the signal diverted. The collected side channel data included photon arrival timing and optical power data obtained using a single-photon detector and a power meter. From this dataset, features capturing both timing dynamics and signal intensity variations were extracted and engineered. Machine learning models were then trained to classify protocol stages based solely on side channel observations. The experimental setup can be represented as:

Why it matters

The study reveals an overlooked vulnerability in quantum identity authentication: side channel leakage can allow an attacker to infer protocol stages and thereby avoid authentication qubits, extracting data qubits and defeating the authentication mechanism. The high accuracy achieved with machine learning models underscores the practical feasibility of such attacks. The authors emphasize the need for robust designs against side channel inference attacks. Potential countermeasures could include reducing side channel leakage, randomizing protocol timing, or employing quantum techniques to detect eavesdropping. The findings call for further research into physical layer security in quantum networks, as side channels may provide a vector for bypassing quantum-layer authentication. The security implication can be modeled as:

where inference accuracy depends on sampling rate and feature quality. This work highlights the importance of considering side channels in the design of quantum authentication protocols.

Who should read this

CS practitioners and researchers

Opening member content…