Original title: Enigma: Privacy-Preserving Execution of QAOA on Untrusted Quantum Computers
Authors: Ramin Ayanzadeh, Ahmad Mousavi, Narges Alavisamani, Moinuddin Qureshi
This article explores privacy concerns when using quantum computers through a method called Enigma. These powerful machines can solve complex problems but are costly and often accessed via external servers, raising worries about data privacy. Existing privacy solutions for quantum computing either demand new tech like quantum networks or create heavy computational burdens. The authors introduce Enigma, a system designed for the Quantum Approximate Optimization Algorithm (QAOA). Unlike other methods that mask quantum circuits, Enigma alters QAOA’s input so the circuit and outcomes stay unreadable to the server. They present three Enigma versions: one secures coefficients, another shields graph nodes using decoy qubits, and the third protects edge information by modifying connections. They validate Enigma’s effectiveness using IBM quantum devices, revealing minimal fidelity loss (1%-13%) while significantly enhancing privacy.
Original article: https://arxiv.org/abs/2311.13546