QCOEM framework boosts quantum cloud fidelity by 30% with zero rescheduling
Achieves 30% higher fidelity and eliminates task rescheduling in noisy quantum clouds.
Quantum cloud platforms struggle to schedule tasks across heterogeneous backends with varying noise profiles and queues. Existing noise-agnostic heuristics cause fidelity loss, load imbalance, and frequent rescheduling.
QCOEM solves this by applying evolutionary algorithms (NSGA-II and NSGA-III) for multi-objective optimization, plus an Augmented Achievement Scalarization Function (AASF) to select Pareto-optimal schedules aligned with user priorities. In simulations, QCOEM achieved 30% higher mean fidelity, zero rescheduling, and maintained low scheduling overhead. This work, accepted at IEEE CLOUD 2026, demonstrates a practical path to stable, high-fidelity quantum cloud computing.
- QCOEM uses NSGA-II and NSGA-III for multi-objective optimization of quantum task scheduling.
- Achieves 30% higher mean fidelity compared to noise-agnostic heuristics.
- Zero task rescheduling and bounded overhead in heterogeneous quantum cloud environments.
Why It Matters
Enables reliable, high-fidelity quantum cloud execution, accelerating practical quantum computing adoption.