Hybrid Optimization Method Using Simulated-Annealing-Based Ising Machine and Quantum Annealer
Shuta Kikuchi, Nozomu Togawa, Shu Tanaka
Abstract
Open-access reader
Shuta Kikuchi, Nozomu Togawa, Shu Tanaka
Abstract
Open-access reader
Ising machines have the potential to realize fast and highly accurate solvers for combinatorial optimization problems. They are classified based on their internal algorithms. Examples include simulated-annealing-based Ising machines (non-quantum-type Ising machines) and quantum-annealing-based Ising machines (quantum annealers). Herein we propose a hybrid optimization method, which utilizes the advantages of both types. In this hybrid optimization method, the preprocessing step is performed by solving the non-quantum-annealing Ising machine multiple times. Then sub-Ising models with a reduced size by spin fixing are solved using a quantum annealer. The performance of the hybrid optimization method is evaluated via simulations using Simulated Annealing (SA) as a non-quantum-type Ising machine and D-Wave Advantage as a quantum annealer. Additionally, we investigate the parameter dependence of the proposed hybrid optimization method. The hybrid optimization method outperforms the preprocessing SA and the quantum annealing machine alone in fully connected random Ising models.
A significance statement is not available in the OpenAlex record.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Ising machines have the potential to realize fast and highly accurate solvers for combinatorial optimization problems. They are classified based on their internal algorithms. Examples include simulated-annealing-based Ising machines (non-quantum-type Ising machines) and quantum-annealing-based Ising machines (quantum annealers). Herein we propose a hybrid optimization method, which utilizes the advantages of both types. In this hybrid optimization method, the preprocessing step is performed by solving the non-quantum-annealing Ising machine multiple times. Then sub-Ising models with a reduced size by spin fixing are solved using a quantum annealer. The performance of the hybrid optimization method is evaluated via simulations using Simulated Annealing (SA) as a non-quantum-type Ising machine and D-Wave Advantage as a quantum annealer. Additionally, we investigate the parameter dependence of the proposed hybrid optimization method. The hybrid optimization method outperforms the preprocessing SA and the quantum annealing machine alone in fully connected random Ising models.
Key concepts: Quantum annealing, Ising model, Quadratic unconstrained binary optimization, Simulated annealing, Quantum, Computer science, Preprocessor, Algorithm