An Optimization Approach for Effective Formalized fUML Model Checking

Abstract

Abstract. Automatically formalizing fUML models into CSP is a challenging task. However, checking the generated CSP model using FDR2 is far more challenging. That is because the generated CSP model holds many implementation details inherited from the fUML model, as well as the formalization of the non-trivial fUML inter-object communication mechanism. Using the state space compression techniques available in FDR2 (such as Supercompilation and compression functions) is not enough to provide an effective model checking that avoids the state explosion problem. In this paper we introduce a novel approach that makes use of a restricted CSP model (because it follows certain formalization rules) to optimize the generated model. As an application of our approach, we design a framework that works on two levels; the first one provides optimization advice to the modeller, while the second one automatically applies optimization rules which transform the CSP model to a more optimized one. Implementing and applying the approach on two large case studies demonstrated the effectiveness of the approach. We also proved that the optimization rules are safe to be applied automatically without eliminating important information from the CSP model.

Decomposition of UML activity diagrams

“…According to the translation rule in Definition 7, the activity diagram  is first translated into the dependency structure td() = ⟨, I, T, S, C, P, F⟩ where  = {0, 1, 2, 3, 4, 5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39, i, f 1, f 2, f 3, f 4, f 5}, I = {{i}}, T = {({i}, {0}), ({0}, {3}), ({0}, {1}), ({1}, {2}), ({2}, {4}), ({2}, {8}), ({4}, {7}), ({7, 8}, {11}), ({3}, {5}), ({5}, {6}), ({6}, {f 1}), ({5}, {9}), ({9}, {10}), ({11}, {12}), ({12}, {13}), ({13}, {f 2}), ({12}, {14}), ({14}, {15}), ({15}, {16}), ({16}, {18}), ({18}, {f 3}), ({16}, {2}), ({15}, {17}), ({17}, {19}), ({10}, {20}), ({19}, {20}), ({20}, {21}), ({20}, {22}), ({21, 22}, {23}), ({23}, {24}), ({24}, {27}), ({27}, {29}),({29}, {30}),({29}, {31}),({30}, {28}),({30}, {25}),({28}, {26}),({26}, {25}), ({25}, {f 4}), ({31}, {20}), ({31}, {32}), ({32}, {33}), ({32}, {36}), ({33}, {34}), ({33}, {35}), ({34, 35}, {37}), ({37}, {38}), ({36}, {39}), ({38}, {39}), ({39}, {f 5})}, S = {{7, 8}, {21, 22}, {34, 35}}, C = {{1, 3}, {6, 9}, {13, 14}, {16, 17}, {25, 28}, {20, 32}, {33, 36}}, P = ∅, and F = {{f 1}, {f 2}, {f 3}, {f 4}, {f 5}}.…”

“…The study revealed interesting insights into the semantic expressivity of activity diagrams and the semantic nature of the different modeling elements, in particular, of the object nodes and the activity final nodes. Abdelhalim et al defined a method of transforming the state machine diagrams and activity diagrams into communicating sequential processes. Later, the method used a model checker to verify the resulting communicating sequential process representation.…”

“…If an activity diagram is mapped to a process, one must add the synchronous event names that express the synchronization controls of the activity diagrams. () Additionally, some existing efforts, eg, Daw and Cleaveland, Grönniger et al, and Knieke have proposed operational semantics for the activity diagrams. Such semantics is actually the transition system‐based structural operational semantics…”