When multiple users share a quantum computer, crosstalk leaks information across tenant boundaries. This project designed an attack that deduces confidential user data by exploiting a crosstalk-based side channel on quantum hardware, using a graph neural network to decipher the victim’s circuit from the information leak — and examined side-channel vulnerabilities in superconducting qubit readout architectures.
Key results#
- Demonstrated circuit-recovery attacks over crosstalk side channels on multi-tenant NISQ computers (NDSS 2025).
- Characterized side-channel vulnerabilities specific to superconducting readout architectures (QCE 2024).
Papers#
Crosstalk-induced Side Channel Threats in Multi-Tenant NISQ Computers Navnil Choudhury, Chaithanya Naik Mude, Sanjay Das, Preetham Tikkireddi, Swamit Tannu, Kanad Basu. NDSS 2025. [Paper]
Understanding Side-Channel Vulnerabilities in Superconducting Qubit Readout Architectures Satvik Maurya, Chaithanya Naik Mude, Benjamin Lienhard, Swamit Tannu. QCE 2024. [Paper]
Code#
Code available upon request — public release coming soon.
BibTeX#
@inproceedings{choudhury2025crosstalk,
title = {Crosstalk-induced Side Channel Threats in Multi-Tenant NISQ Computers},
author = {Choudhury, Navnil and Mude, Chaithanya Naik and Das, Sanjay and Tikkireddi, Preetham and Tannu, Swamit and Basu, Kanad},
booktitle = {Network and Distributed System Security Symposium (NDSS)},
year = {2025}
}
@inproceedings{maurya2024sidechannel,
title = {Understanding Side-Channel Vulnerabilities in Superconducting Qubit Readout Architectures},
author = {Maurya, Satvik and Mude, Chaithanya Naik and Lienhard, Benjamin and Tannu, Swamit},
booktitle = {IEEE International Conference on Quantum Computing and Engineering (QCE)},
year = {2024}
}