Confidential QUBO solver

Abstract

Quadratic Unconstrained Binary Optimization (QUBO) is widespread and solvable via classical or quantum computing. However, outsourcing these computations online exposes sensitive data to potential breaches. We introduce a novel encryption scheme that seamlessly integrates with any solver or hardware platform, ensuring data security without compromising performance. A robust python implementation delivers promising results, marking a significant step forward in secure optimization for both classical and quantum environments.