Géza Kulcsár

Topology control with application constraints

Numerous topology control algorithms for wireless sensor networks exist. Typically, these algorithms optimize general network metrics like the transmission range of sensors. Being of general nature, they suit application-specific requirements insufficiently; e.g., application-specific communication patterns may favor links that the general algorithm will remove. We suggest a new research focus called application-specific topology control and propose a methodic approach for which we provide a first exploration. First, application-specific communication patterns are expressed as overlay graphs. Then, application-specific requirements are specified as constraints on this overlay. Next follows the core of the method, geared toward deriving application-specific topology control algorithms. Applying the method to a data collection application with a many-to-one communication pattern, we propose two topology control algorithms. In a simulation study, we evaluate these algorithms against the existing algorithm kTC and a shortest path tree, showing that the method allows trading off general network optimization against application-specific optimization.

Improved Conflict Detection for Graph Transformation with Attributes

In graph transformation, a conflict describes a situation where two alternative transformations cannot be arbitrarily serialized. When enriching graphs with attributes, existing conflict detection techniques typically report a conflict whenever at least one of two transformations manipulates a shared attribute. In this paper, we propose an improved, less conservative condition for static conflict detection of graph transformation with attributes by explicitly taking the semantics of the attribute operations into account. The proposed technique is based on symbolic graphs, which extend the traditional notion of graphs by logic formulas used for attribute handling. The approach is proven complete, i.e., any potential conflict is guaranteed to be detected.

Rapid Prototyping of Topology Control Algorithms by Graph Transformation

Topology control algorithms are used to improve the energy efficiency (or other quality parameters) of wireless sensor networks. In this paper, we propose a model-driven rapid prototyping approach for the kTC topology control algorithm to enable the fast implementation and the evaluation of its different variants, and consequently, to accelerate the network quality experimentation cycle. In our approach, wireless sensor networks are described by graph-based models, and three variants of the kTC topology control algorithm are implemented by graph transformation, which are then executed on input network descriptions to derive modified topologies whose quality is then measured in several contexts to be able to assess the achieved network quality improvement.

Continuous detection of design flaws in evolving object-oriented programs using incremental multi-pattern matching

Design flaws in object-oriented programs may seriously corrupt code quality thus increasing the risk for introducing subtle errors during software maintenance and evolution. Most recent approaches identify design flaws in an ad-hoc manner, either focusing on software metrics, locally restricted code smells, or on coarse-grained architectural anti-patterns. In this paper, we utilize an abstract program model capturing high-level object-oriented code entities, further augmented with qualitative and quantitative design-related information such as coupling/cohesion. Based on this model, we propose a comprehensive methodology for specifying object-oriented design flaws by means of compound rules integrating code metrics, code smells and anti-patterns in a modular way. This approach allows for efficient, automated design-flaw detection through incremental multi-pattern matching, by facilitating systematic information reuse among multiple detection rules as well as between subsequent detection runs on continuously evolving programs. Our tool implementation comprises well-known anti-patterns for Java programs. The results of our experimental evaluation show high detection precision, scalability to real-size programs, as well as a remarkable gain in efficiency due to information reuse.

Conflict Detection for Edits on Extended Feature Models using Symbolic Graph Transformation

Feature models are used to specify variability of user-configurable systems as appearing, e.g., in software product lines. Software product lines are supposed to be long-living and, therefore, have to continuously evolve over time to meet ever-changing requirements. Evolution imposes changes to feature models in terms of edit operations. Ensuring consistency of concurrent edits requires appropriate conflict detection techniques. However, recent approaches fail to handle crucial subtleties of extended feature models, namely constraints mixing feature-tree patterns with first-order logic formulas over non-Boolean feature attributes with potentially infinite value domains. In this paper, we propose a novel conflict detection approach based on symbolic graph transformation to facilitate concurrent edits on extended feature models. We describe extended feature models formally with symbolic graphs and edit operations with symbolic graph transformation rules combining graph patterns with first-order logic formulas. The approach is implemented by combining eMoflon with an SMT solver, and evaluated with respect to applicability.

Incremental Co-Evolution of Java Programs based on Bidirectional Graph Transformation

Modern Java IDE aim at assisting object-oriented software development workflows with continuously interleaved co-evolution steps of program editing and program refactoring. Program editing usually comprises manually performed program changes applied by a programmer at source code level. In contrast, refactorings consist of behavior-preserving program restructuring rules with complex preconditions, usually formulated over an appropriate program abstraction. To integrate both steps into a comprehensive program evolution framework, we present a graph-based approach for incremental co-evolution of Java programs. Our approach is based on a concise graph-based representation of Java programs by means of a reduced abstract syntax tree, augmented with additional cross-tree edges denoting crucial semantic information. On this basis, a precise formal specification of object-oriented program refactorings can be defined in terms of endogenous graph-transformation rules. In addition, we use Triple Graph Grammars (TGG) to define exogenous bidirectional graph transformation rules for automated incremental synchronization between a program graph and the corresponding source code. Our implementation relies on the graph-transformation engine eMoflon and currently supports the Java refactorings Pull Up Method and Create Superclass.

Equivalence and Independence in Controlled Graph-Rewriting Processes

Graph transformation systems (GTS) are often defined as sets of rules that can be applied repeatedly and non-deterministically to model the evolution of a system. Several semantics proposed for GTSs are relevant in this case, providing means for analysing the system’s behaviour in terms of dependencies, conflicts and potential parallelism among the relevant events. Several other approaches equip GTSs with an additional control layer useful for specifying rule application strategies, for example to describe graph manipulation algorithms. Almost invariably, the latter approaches consider only an input-output semantics, for which the above mentioned semantics are irrelevant.

Graph-Rewriting Petri Nets.

Controlled graph rewriting enhances expressiveness of plain graph-rewriting systems (i.e., sets of graph-rewriting rules) by introducing additional constructs for explicitly controlling graph-rewriting rule applications. In this regard, a formal semantic foundation for controlled graph rewriting is inevitable as a reliable basis for tool-based specification and automated analysis of graph-based algorithms. Although several promising attempts have been proposed in the literature, a comprehensive theory of controlled graph rewriting capturing semantic subtleties of advanced control constructs provided by practical tools is still an open challenge. In this paper, we propose graph-rewriting Petri nets (GPN) as a novel foundation for unifying control-flow and rule-application semantics of controlled graph rewriting. GPN instantiate coloured Petri nets with categorical DPO-based graph-rewriting theory where token colours denote typed graphs and graph morphisms and transitions define templates for guarded graph-rewriting rule applications. Hence, GPN enjoy the rich body of specification and analysis techniques of Petri nets including inherent notions of concurrency. To demonstrate expressiveness of GPN, we present a case study by means of a topology-control algorithm for wireless sensor networks.

Controlling the Attack Surface of Object-Oriented Refactorings

Refactorings constitute an effective means to improve quality and maintainability of evolving object-oriented programs. Search-based techniques have shown promising results in finding optimal sequences of behavior-preserving program transformations that (1) maximize code-quality metrics and (2) minimize the number of changes. However, the impact of refactorings on extra-functional properties like security has received little attention so far. To this end, we propose as a further objective to minimize the attack surface of programs (i.e., to maximize strictness of declared accessibility of class members). Minimizing the attack surface naturally competes with applicability of established MoveMethod refactorings for improving coupling/cohesion metrics. Our tool implementation is based on an EMF meta-model for Java-like programs and utilizes MOMoT, a search-based model-transformation framework. Our experimental results gained from a collection of real-world Java programs show the impact of attack surface minimization on design-improving refactorings by using different accessibility-control strategies. We further compare the results to those of existing refactoring tools.

Towards a Step Semantics for Story-Driven Modelling

Graph Transformation (GraTra) provides a formal, declarative means of specifying model transformation. In practice, GraTra rule applications are often programmed via an additional language with which the order of rule applications can be suitably controlled. Story-Driven Modelling (SDM) is a dialect of programmed GraTra, originally developed as part of the Fujaba CASE tool suite. Using an intuitive, UML-inspired visual syntax, SDM provides usual imperative control flow constructs such as sequences, conditionals and loops that are fairly simple, but whose interaction with individual GraTra rules is nonetheless non-trivial. In this paper, we present the first results of our ongoing work towards providing a formal step semantics for SDM, which focuses on the execution of an SDM specification.