Pattern-Based Generation and Adaptation of Quantum Workflows

Building quantum applications requires deep knowledge of quantum computing and software engineering. Hence, an abstraction layer reducing the complexity for non-experts is needed. Patterns are an established concept for the abstract description of proven solutions to recurring problems. Therefore, t...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Proceedings / International Conference on Software Engineering s. 3072 - 3084
Hlavní autoři: Beisel, Martin, Barzen, Johanna, Leymann, Frank, Stiliadou, Lavinia, Vietz, Daniel, Weder, Benjamin
Médium: Konferenční příspěvek
Jazyk:angličtina
Vydáno: IEEE 26.04.2025
Témata:
Adaptation models
Complexity theory
Computational modeling
Computer architecture
Knowledge engineering
Patterns
Prototypes
Quantum computing
Runtime
Software
Workflows
ISSN:1558-1225
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:Building quantum applications requires deep knowledge of quantum computing and software engineering. Hence, an abstraction layer reducing the complexity for non-experts is needed. Patterns are an established concept for the abstract description of proven solutions to recurring problems. Therefore, the quantum computing patterns, a pattern language for the quantum computing domain, can be used to define the building blocks and the structure of hybrid quantum applications. Furthermore, concrete software artifacts can be associated with patterns to solve the corresponding problem. However, these software artifacts are usually heterogeneous, e.g., using different data formats. Quantum workflows enable a robust and scalable orchestration of these heterogeneous software artifacts. However, manually modeling and configuring such quantum workflows is a complex, error-prone, and time-consuming task. To overcome this issue, we present an approach that automates the generation and adaptation of quantum workflows using the quantum computing patterns. We provide an architecture realizing our approach, a corresponding prototype, as well as an evaluation comprising different use cases, a runtime comparison, and a user study.
ISSN:1558-1225
DOI:10.1109/ICSE55347.2025.00196