Optimized continuous dynamical decoupling via differential geometry and machine learning
| Autor: | Morazotti, Nicolas André da Costa, da Silva, Adonai Hilário, Audi, Gabriel, Fanchini, Felipe Fernandes, Napolitano, Reginaldo de Jesus |
|---|---|
| Rok vydání: | 2023 |
| Předmět: | |
| Zdroj: | Phys. Rev. A 110, 042601 (2024) |
| Druh dokumentu: | Working Paper |
| DOI: | 10.1103/PhysRevA.110.042601 |
| Popis: | We introduce a strategy to develop optimally designed fields for continuous dynamical decoupling. Using our methodology, we obtain the optimal continuous field configuration to maximize the fidelity of a general one-qubit quantum gate. To achieve this, considering dephasing-noise perturbations, we employ an auxiliary qubit instead of the boson bath to implement a purification scheme, which results in unitary dynamics. Employing the sub-Riemannian geometry framework for the two-qubit unitary group, we derive and numerically solve the geodesic equations, obtaining the optimal time-dependent control Hamiltonian. Also, due to the extended time required to find solutions to the geodesic equations, we train a neural network on a subset of geodesic solutions, enabling us to promptly generate the time-dependent control Hamiltonian for any desired gate, which is crucial in circuit optimization. Comment: 14 pages, 10 figures |
| Databáze: | arXiv |
| Externí odkaz: |
|