The use of distributed systems, involving multiple components, has become a common industry practice. However, modeling the behaviour of such systems is a challenge, especially when the behavior consists of several collaborations of different parties, each involving possibly several starting (input) and ending (output) events of the involved components. Furthermore, the global behavior should be described as a composition of several sub-behaviours, in the following called collaborations, and each collaboration may be further decomposed into several sub-collaborations. We assume that the performance of the elementary sub-collaborations is known, and that the performance of the global behavior should be determined from the performance of the contained elementary collaborations and the form of the composition. A collaboration, in this thesis, is characterized by a partial order of input and output events, and the performance of the collaboration is defined by the minimum delays required for a given output event with respect to an input event. This is a generalization of the semantics of UML Activities, where all input events are assumed to occur at the same time, and all output events occur at the same time. We give a semantic definition of the dynamic behavior of composed collaborations using the composition operators for control flow from UML Activity diagrams, in terms of partial order relationships among the involved input and output events. Based on these semantics, we provide formulas for calculating the performance of composed collaborations in terms of the performance of the sub-collaborations, where each delay is characterized by (a) a fixed value, (b) a range of values, and (c) a distribution (in the case of stochastic behaviours). We also propose approximations for the case of stochastic behavior with Normal distributions, and discuss the expected errors that may be introduced due to ignoring of shared resources or possible dependencies in the case of stochastic behaviours. A tool has been developed for evaluating the performance of complex collaborations, and examples and case studies are discussed to illustrate the applicability of the performance analysis and the visual notation which we introduced for representing the partial-order relationships of the input and output events.

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / This work presents a method for qualitative and quantitative analysis of WorkFlow nets based on the proof trees of linear logic, and an approach for the verification of workflow specifications in UML through the transformation of UML Activity Diagrams into WorkFlow nets. The qualitative analysis is concerned with the proof of soundness correctness criterion defined for WorkFlow nets. The quantitative analysis is based on the computation of symbolic dates for the planning of resources used to handle each task of the workflow process modeled by a t-Time WorkFlow net. For the verification of the specifications of workflow processes mapped into UML Activity Diagrams are presented formal rules to transform this ones into WorkFlow nets. In this context is proposed the analysis and correction of critical points in UML Activity Diagrams through the analysis of proof trees of linear logic. The advantages of such an approach are diverse. The fact of working with linear logic permits one to prove the correctness criterion soundness in a linear time without considering the construction of the reachability graph, considering the proper structure of the WorkFlow net instead of considering the corresponding automata. Moreover, the computation of symbolic dates for the execution of each task mapped into the t-Time WorkFlow net permits to plan the utilization of the resources involved in the activities of the workflow process, through formulas that can be used for any case handled by the correspondent workflow process, without to examine again the process to recalculate, for each new case, the dates of start and conclusion for the activities involved in the process. Regarding the verification of workflow processes mapped into UML Activity Diagrams, the major advantage of this approach is the transformation of a semi-formal model into a formal model, such that some properties, like soundness, can be formally verified. / Este trabalho apresenta um método para a análise qualitativa e quantitativa de Work- Flow nets baseado nas árvores de prova canônica da lógica linear e uma abordagem para a verificação de especificações de processos de workflow em UML através da transformação de Diagramas de Atividades da UML em WorkFlow nets. A análise qualitativa refere-se à prova do critério de corretude soundness definido para WorkFlow nets. Já a análise quantitativa preocupa-se com o planejamento de recursos para cada atividade de um processo de workflow mapeado em uma t-Time WorkFlow net e baseia-se no cálculo de datas simbólicas para o planejamento de recursos utilizados na realização de cada tarefa do processo de workflow. Para a verificação das especificações de processos de workflow mapeados em Diagramas de Atividades da UML são apresentadas regras formais para transformar estes diagramas em WorkFlow nets. Neste contexto também é proposta a análise e correção de pontos críticos em Diagramas de Atividades da UML através da análise de árvores de prova canônica da lógica linear. As vantagens das abordagens apresentadas neste trabalho são diversas. O fato de trabalhar com lógica linear permite provar o critério de corretude soundness em tempo linear e sem que seja necessária a construção de um grafo das marcações acessíveis, considerando diretamente a própria estrutura da WorkFlow net, ao invés de considerar o seu autômato correspondente. Além disso, o cálculo de datas simbólicas correspondentes à execução de cada tarefa mapeada em uma t-Time WorkFlow net permite planejar a utilização dos recursos envolvidos nas atividades do processo de workflow, através de fórmulas que podem ser utilizadas por qualquer caso tratado pelo processo de workflow correspondente, sem que seja necessário percorrer novamente o processo de workflow inteiro para recalcular, para cada novo caso, datas de início e término das atividades envolvidas no processo. Já no que diz respeito à verificação de processos de workflow mapeados em Diagramas de Atividades da UML, a principal vantagem desta abordagem é a transformação de um modelo semi-formal em um modelo formal, para o qual algumas propriedades, como soundness, podem ser formalmente verificadas. / Mestre em Ciência da Computação

The use of distributed systems, involving multiple components, has become a common industry practice. However, modeling the behaviour of such systems is a challenge, especially when the behavior consists of several collaborations of different parties, each involving possibly several starting (input) and ending (output) events of the involved components. Furthermore, the global behavior should be described as a composition of several sub-behaviours, in the following called collaborations, and each collaboration may be further decomposed into several sub-collaborations. We assume that the performance of the elementary sub-collaborations is known, and that the performance of the global behavior should be determined from the performance of the contained elementary collaborations and the form of the composition. A collaboration, in this thesis, is characterized by a partial order of input and output events, and the performance of the collaboration is defined by the minimum delays required for a given output event with respect to an input event. This is a generalization of the semantics of UML Activities, where all input events are assumed to occur at the same time, and all output events occur at the same time. We give a semantic definition of the dynamic behavior of composed collaborations using the composition operators for control flow from UML Activity diagrams, in terms of partial order relationships among the involved input and output events. Based on these semantics, we provide formulas for calculating the performance of composed collaborations in terms of the performance of the sub-collaborations, where each delay is characterized by (a) a fixed value, (b) a range of values, and (c) a distribution (in the case of stochastic behaviours). We also propose approximations for the case of stochastic behavior with Normal distributions, and discuss the expected errors that may be introduced due to ignoring of shared resources or possible dependencies in the case of stochastic behaviours. A tool has been developed for evaluating the performance of complex collaborations, and examples and case studies are discussed to illustrate the applicability of the performance analysis and the visual notation which we introduced for representing the partial-order relationships of the input and output events.