quantum-classical optimization

Quantum-classical hybrid programs lack systematic compilation optimization methodologies and standardized metrics for evaluating co-execution efficiency. Method: This paper proposes the first compilation-level optimization framework targeting real-time quantum-classical co-computation. It introduces seven compiler optimizations—including quantum-classical communication cost modeling, quantum circuit simplification, and classical control-flow restructuring—and defines three quantitative co-execution efficiency metrics. An end-to-end optimizing compiler is implemented atop the Quil instruction language. Results: Experiments across multiple hybrid benchmarks demonstrate substantial reductions in quantum-classical communication rounds and total execution latency, achieving an average 32.7% improvement in end-to-end co-execution efficiency. Contribution: This work establishes the first compilation optimization paradigm and dedicated evaluation framework for hybrid quantum-classical programs, providing a scalable compiler infrastructure to support real-time quantum-classical co-processing.

The growing complexity of software engineering (SE) optimization problems necessitates rigorous investigation into emerging paradigms such as quantum computing. Method: We conducted the first systematic literature review (SLR) specifically focused on quantum optimization in SE, screening 2,083 publications from six major academic databases and selecting 77 empirical studies through structured searching, multi-stage filtering, and thematic coding. Contribution/Results: Our analysis constructs the first comprehensive research landscape of quantum optimization in SE, revealing strong concentration in test case generation and deployment optimization, while critical areas—including requirements engineering, maintenance, and software evolution—remain severely underexplored. Furthermore, many influential contributions appear outside mainstream SE venues, highlighting pronounced interdisciplinary fragmentation. The study identifies key research gaps and proposes concrete collaborative pathways, thereby establishing both a theoretical foundation and practical guidance for advancing SE–quantum computing convergence.

This study addresses industrial job shop scheduling by developing a customized model that incorporates hardware constraints and systematically evaluates the performance of quantum annealing (D-Wave), digital annealing (Fujitsu), quantum-inspired algorithms, and classical approaches—including mixed-integer linear programming (MILP) and exact solvers—on platforms such as IBM Quantum. Through a hardware-software co-design methodology, the work demonstrates the practical utility of quantum and quantum-inspired techniques in enhancing the quality of approximate solutions, guiding solver selection, and integrating into classical computational workflows. The research establishes a scalable hybrid solving paradigm for industrial scheduling and validates its feasibility and potential during early-stage proof-of-concept demonstrations.

Quantum-classical hybrid solvers face integration bottlenecks in existing optimization workflows due to the absence of a unified software stack. Method: This paper proposes Meta-Solver—a meta-solution framework built on a unified software architecture that enables problem-adaptive decomposition and coordinated scheduling of classical and quantum subtasks. It integrates mainstream classical optimization algorithms with configurable quantum circuits, supports multiple quantum hardware backends, and provides an interactive configuration interface. Contribution/Results: Evaluated on real-world applications, Meta-Solver significantly improves plug-and-play capability and deployment efficiency of quantum subroutines. It represents the first lightweight, scalable hybrid solving paradigm tailored for industrial optimization pipelines, establishing critical infrastructure for the incremental integration of quantum advantage into practical systems.

During the Qiskit 1.x→2.x migration, the default shot count for QAOA was drastically reduced, causing insufficient state-space coverage (only 23%) and leading to output distribution shifts and significant accuracy degradation—undermining reproducibility. Method: We developed a standardized QAOA implementation based on Qiskit 2.x v2 primitives, rigorously controlling circuit construction, optimizer selection, and Hamiltonian encoding, and systematically quantified performance decay across varying shot counts. Contribution/Results: We identified the shot reduction as the root cause and proposed a reproducibility-preserving shot count of 250,000, which fully restores original accuracy. This work is the first to expose the critical impact of implicit parameters in the quantum-classical interface layer on hybrid algorithm performance. It establishes a parameter calibration paradigm and empirical benchmark for quantum software version migration, enabling robust cross-version algorithm deployment.

This work addresses the challenge of meeting user quality-of-service (QoS) requirements under the constraints of noisy intermediate-scale quantum (NISQ) devices by proposing a service-oriented architecture for hybrid quantum-classical systems. It pioneers the integration of service-oriented architecture (SOA) with quantum computing, employing formal architectural style modeling and QoS-driven design space exploration to delineate architectural decision boundaries. The approach dynamically selects optimal execution strategies at both structural and behavioral levels in response to varying QoS demands. Experimental results demonstrate that the proposed method can dynamically configure the system under realistic NISQ constraints to deliver quantifiable performance guarantees aligned with user-specified QoS requirements.

This work addresses the challenge of identifying suitable quantum computing candidates and expressing solution strategies within hybrid quantum-classical workflows, which is often hindered by highly specialized expertise and a lack of unified abstractions. To overcome these limitations, the paper introduces the first domain-specific language (DSL) tailored for hybrid quantum-classical meta-solving strategies. This DSL enables technology-agnostic, problem-level specification of strategies and is accompanied by a dedicated execution framework. By integrating with the ProvideQ toolkit for workflow orchestration, the proposed open-source solution significantly lowers the barrier to entry, enhances strategy reusability, and improves automation—thereby facilitating the identification and deployment of quantum-suitable problems.

Current quantum software development overly emphasizes algorithms while neglecting architectural design, thereby limiting the scalability and engineering maturity of hybrid quantum-classical systems. To address this gap, this work proposes the Quantum Software Architecture Framework (QSAF), which for the first time categorizes 34 quantum circuit primitives by functionality and abstracts them into reusable architectural components with explicit interfaces and design constraints. QSAF establishes a multi-level abstraction hierarchy spanning from quantum gates to system-level constructs. By incorporating non-functional properties—such as circuit depth, error sensitivity, and information flow—to characterize component behavior, QSAF enables structured decomposition and optimization of hybrid workflows like variational quantum algorithms, significantly enhancing the rigor, modularity, and development efficiency of quantum system design.

This work addresses the lack of reusable, multi-QPU-cooperative simulation tools for the Quantum Approximate Optimization Algorithm (QAOA) in engineering design and decision-making involving Quadratic Unconstrained Binary Optimization (QUBO) problems. We present the first distributed QAOA simulation framework that supports user-defined numbers and capacities of quantum processing units (QPUs). Built on Qiskit, the framework fully integrates QUBo modeling, distributed variable allocation, cross-QPU coupling handling, parameterized quantum circuit generation, and a Streamlit-based graphical interface. Runtime optimizations—including circuit reuse, batched evaluation, and parallel multi-start strategies—are incorporated to enhance efficiency. Experiments on standard QUBO benchmarks and the unit commitment problem demonstrate that both distributed and monolithic QAOA implementations recover optimal solutions, with staged optimization significantly reducing runtime while maintaining consistency with classical single-QPU QAOA results.

Quantum software and artificial intelligence development face significant challenges, including a scarcity of skilled personnel, low development efficiency, and complex deployment decisions in hybrid systems. This study presents the first systematic literature review on automated software engineering and AI methods specifically tailored for quantum and hybrid quantum-classical systems. It synthesizes existing techniques, tools, and application strategies, while identifying key automation approaches and critical research gaps. By addressing the lack of comprehensive reviews in this interdisciplinary domain, the work establishes a theoretical foundation and offers practical pathways to enhance the development efficiency and deployment intelligence of quantum–AI integrated systems.