Universität Stuttgart

Permanent URI for this communityhttps://elib.uni-stuttgart.de/handle/11682/1

Browse

Filters

Settings

Institute: search.filters.UBSInstitut.Institut für Architektur von AnwendungssystemenAuthor: search.filters.author.Leymann, FrankSubject: search.filters.subject.004

Search Results

Now showing 1 - 10 of 22
  • Thumbnail Image
    ItemOpen Access
    Abstract syntax of WS-BPEL 2.0
    (2008) Kopp, Oliver; Mietzner, Ralph; Leymann, Frank
    WS-BPEL 2.0 is the current version of the "Business Process Execution Language for Web Services". Until now, no formal definition of its syntax exists. We present a complete syntax of WS-BPEL 2.0 of both abstract and executable processes.
  • Thumbnail Image
    ItemOpen Access
    Propagation of states from BPEL process instances to Chevron models
    (2011) Schumm, David; Karastoyanova, Dimka; Leymann, Frank; Lie, Sumadi
    This report describes key aspects of a code library that we developed for the purpose of state propagation for business process monitoring on different levels of abstraction. The library supports the propagation of execution states of process instances based on the Business Process Execution Language (BPEL) to process models specified in the "Chevron" language. The Chevron language is an abstract, non-executable process language that we especially designed for abstract process instance monitoring purposes. The look and feel of this graphical language is similar to value chains. The basic concept of the Chevron language is based on Chevron-shaped charts which can be modeled in Microsoft PowerPoint to describe a process on a high level of abstraction. We aim at enabling the use of high-level process in order to monitor the instance status of a much more detailed, lower-level model. We describe the overall procedure of performing state projections along a concrete scenario. We describe a format for state propagation rules which define how the status of activities of a BPEL process instance should be projected to the elements of a Chevron model. We present a format to serialize process models in the Chevron language. We present a graphical template based on Scalable Vector Graphics (SVG) which we employ to render a stateful Chevron model graphically. The Chevron language is just one language to be used for abstract representation of process instances. However, the approach for state propagation is generic and can be applied for other languages, too.
  • Thumbnail Image
    ItemOpen Access
    On reducing the amount of samples required for training of QNNs : constraints on the linear structure of the training data
    (2025) Mandl, Alexander; Barzen, Johanna; Leymann, Frank; Vietz, Daniel
    Training classical neural networks generally requires a large number of training samples. Using entangled training samples, Quantum Neural Networks (QNNs) have the potential to significantly reduce the amount of training samples required in the training process. However, to minimize the number of incorrect predictions made by the resulting QNN, it is essential that the structure of the training samples meets certain requirements. On the one hand, the exact degree of entanglement must be fixed for the whole set of training samples. On the other hand, training samples must be linearly independent and non-orthogonal. However, how failing to meet these requirements affects the resulting QNN is not fully studied. To address this, we extend the proof of the Quantum No-Free-Lunch theorem to (i) provide a generalization of the theorem for varying degrees of entanglement. This generalization shows that the average degree of entanglement in the set of training samples can be used to predict the expected quality of the QNN. Furthermore, we (ii) introduce new estimates for the expected accuracy of QNNs for moderately entangled training samples that are linearly dependent or orthogonal. Our analytical results are (iii) experimentally validated by simulating QNN training and analyzing the quality of the QNN after training.
  • Thumbnail Image
    ItemOpen Access
    Coordination protocols for split BPEL loops and scopes
    (2007) Khalaf, Rania; Leymann, Frank
    The document presents an approach to enable loops and fault handling, compensating scopes to be split among a set of BPEL processes running on different BPEL engines. A mechanism to split a scope or loop into multiple fragments is presented, then a protocol is defined that can be used to coordinate fragments of a loop or a scope so that those fragments run as if they had been in a single process. The requirements for running split scopes and loops are explained. For compensation, this paper focuses on explicit compensation and makes the assumption that compensation handing does not fail. Two protocols are defined such that they may be plugged into the WS-Coordination framework. The messages between the participant fragments and the coordinator are defined. The information about the participating processes that the coordinator needs to have is specified. An algorithm is provided to locate a fault handler in the hierarchy of scopes that can handle a particular BPEL fault. Additionally, the behavior of both participants and the coordinator are specified.
  • Thumbnail Image
    ItemOpen Access
    Automated quantum hardware selection for quantum workflows
    (2021) Weder, Benjamin; Barzen, Johanna; Leymann, Frank; Salm, Marie
    The execution of a quantum algorithm typically requires various classical pre- and post-processing tasks. Hence, workflows are a promising means to orchestrate these tasks, benefiting from their reliability, robustness, and features, such as transactional processing. However, the implementations of the tasks may be very heterogeneous and they depend on the quantum hardware used to execute the quantum circuits of the algorithm. Additionally, today’s quantum computers are still restricted, which limits the size of the quantum circuits that can be executed. As the circuit size often depends on the input data of the algorithm, the selection of quantum hardware to execute a quantum circuit must be done at workflow runtime. However, modeling all possible alternative tasks would clutter the workflow model and require its adaptation whenever a new quantum computer or software tool is released. To overcome this problem, we introduce an approach to automatically select suitable quantum hardware for the execution of quantum circuits in workflows. Furthermore, it enables the dynamic adaptation of the workflows, depending on the selection at runtime based on reusable workflow fragments. We validate our approach with a prototypical implementation and a case study demonstrating the hardware selection for Simon’s algorithm.
  • Thumbnail Image
    ItemOpen Access
    The bitter truth about gate-based quantum algorithms in the NISQ era
    (2020) Leymann, Frank; Barzen, Johanna
    Implementing a gate-based quantum algorithm on an noisy intermediate scale quantum (NISQ) device has several challenges that arise from the fact that such devices are noisy and have limited quantum resources. Thus, various factors contributing to the depth and width as well as to the noise of an implementation of a gate-based algorithm must be understood in order to assess whether an implementation will execute successfully on a given NISQ device. In this contribution, we discuss these factors and their impact on algorithm implementations. Especially, we will cover state preparation, oracle expansion, connectivity, circuit rewriting, and readout: these factors are very often ignored when presenting a gate-based algorithm but they are crucial when implementing such an algorithm on near-term quantum computers. Our contribution will help developers in charge of realizing gate-based algorithms on such machines in (i) achieving an executable implementation, and (ii) assessing the success of their implementation on a given machine.
  • Thumbnail Image
    ItemOpen Access
    Configurable readout error mitigation in quantum workflows
    (2022) Beisel, Martin; Barzen, Johanna; Leymann, Frank; Truger, Felix; Weder, Benjamin; Yussupov, Vladimir
    Current quantum computers are still error-prone, with measurement errors being one of the factors limiting the scalability of quantum devices. To reduce their impact, a variety of readout error mitigation methods, mostly relying on classical post-processing, have been developed. However, the application of these methods is complicated by their heterogeneity and a lack of information regarding their functionality, configuration, and integration. To facilitate their use, we provide an overview of existing methods, and evaluate general and method-specific configuration options. Quantum applications comprise many classical pre- and post-processing tasks, including readout error mitigation. Automation can facilitate the execution of these often complex tasks, as their manual execution is time-consuming and error-prone. Workflow technology is a promising candidate for the orchestration of heterogeneous tasks, offering advantages such as reliability, robustness, and monitoring capabilities. In this paper, we present an approach to abstractly model quantum workflows comprising configurable readout error mitigation tasks. Based on the method configuration, these workflows can then be automatically refined into executable workflow models. To validate the feasibility of our approach, we provide a prototypical implementation and demonstrate it in a case study from the quantum humanities domain.
  • Thumbnail Image
    ItemOpen Access
    Loss behavior in supervised learning with entangled states
    (2026) Mandl, Alexander; Barzen, Johanna; Bechtold, Marvin; Leymann, Frank; Stiliadou, Lavinia
    Quantum machine learning (QML) aims to leverage the principles of quantum mechanics to speed up the process of solving machine learning problems or improve the quality of solutions. In supervised learning, training samples that are entangled with an auxiliary system were shown to reduce the approximation error of the trained model. However, also the training process is affected by various properties of the supervised learning task, including the model's circuit structure, the used cost function, and noise on the quantum computer. To evaluate the applicability of entanglement in supervised learning, the effect of highly entangled training data on the trainability is investigated in this work. This study shows that for highly expressive models, i.e., models capable of expressing a large number of candidate solutions, the possible improvement of loss function values in constrained neighborhoods during optimization is severely limited when maximally entangled states are used. Furthermore, this finding is experimentally supported by simulating training with Parameterized Quantum Circuits (PQCs). As the expressivity of the PQC increases, this training process becomes more susceptible to loss concentration induced by entangled training data. Lastly, for non‐maximally entangled states, the experiments highlight the fundamental role of entanglement entropy as a predictor for the trainability.
  • Thumbnail Image
    ItemOpen Access
    Provenance-preserving analysis and rewrite of quantum workflows for hybrid quantum algorithms
    (2023) Weder, Benjamin; Barzen, Johanna; Beisel, Martin; Leymann, Frank
    Quantum applications are hybrid, i.e., they comprise quantum and classical programs, which must be orchestrated. Workflows are a proven solution for orchestrating heterogeneous programs while providing benefits, such as robustness or scalability. However, the orchestration using workflows can be inefficient for some quantum algorithms, requiring the execution of quantum and classical programs in a loop. Hybrid runtimes are offered to efficiently execute these algorithms. For this, the quantum and classical programs are combined in a single hybrid program, for which the execution is optimized. However, this leads to a conceptual gap between the modeling benefits of workflow technologies, e.g., modularization, reuse, and understandability, and the efficiency improvements when using hybrid runtimes. To close this gap, we introduce a method to model all tasks explicitly in the workflow model and analyze the workflow to detect parts of the workflow that can benefit from hybrid runtimes. Furthermore, corresponding hybrid programs are automatically generated based on the quantum and classical programs, and the workflow is rewritten to invoke them. To ease the live monitoring and later analysis of workflow executions, we integrate process views into our method and collect related provenance data. Thus, the user can visualize and monitor the workflow in the original and rewritten form within the workflow engine. The practical feasibility of our approach is validated by a prototypical implementation, a case study, and a runtime evaluation.
  • Thumbnail Image
    ItemOpen Access
    WS-BPEL extension for semantic web services (BPEL4SWS), version 1.0
    (2008) Karastoyanova, Dimka; Van Lessen, Tammo; Leymann, Frank; Nitzsche, Joerg; Wutke, Daniel
    The Web Services Business Process Execution Language, version 2.0 (WS-BPEL 2.0 or BPEL for brevity) introduces a model for business processes based on Web services. A BPEL process orchestrates interactions among different Web services. The language encompasses features needed to describe complex control flows, including error handling and compensation behavior. BPEL for Semantic Web Services (BPEL4SWS) uses Semantic Web Service Frameworks to define a communication channel between two partner services instead of using the partner link which is based on WSDL 1.1. It enables describing activity implementations in a much more flexible manner based on ontological descriptions of service requesters and providers.