@phdthesis{Beyer2002, author = {Beyer, Dirk}, title = {Formale Verifikation von Realzeit-Systemen mittels Cottbus Timed Automata}, url = {http://nbn-resolving.de/urn:nbn:de:kobv:co1-000000258}, school = {BTU Cottbus - Senftenberg}, year = {2002}, abstract = {Die Konstruktion eingebetteter Systeme, die starke Realzeit-Anforderungen zu erf{\"u}llen haben, wird in den verschiedensten Anwendungsbereichen immer bedeutsamer, z. B. in der Medizin, der Transporttechnik oder der Produktionsautomatisierung. Formale Methoden unterst{\"u}tzen die fehlerarme Entwicklung solcher Systeme, weil sie auf einer pr{\"a}zisen mathematischen Grundlage aufbauen. Der Autor entwickelt einen geeigneten Modellierungsformalismus und effiziente Verifikationsverfahren f{\"u}r den Einsatz einer formalen Methode. Durch die Einf{\"u}hrung eines Modulkonzepts wird die Modellierung auch großer Realzeit-Systeme systematisch unterst{\"u}tzt. F{\"u}r die Verifikation werden effiziente BDD-basierte Algorithmen verwendet, wobei auch das Problem des Findens guter BDD-Variablenordnungen gel{\"o}st wird. Es werden sowohl Erreichbarkeitsanalyse als auch Verfeinerungsanalyse unterst{\"u}tzt. Die Praktikabilit{\"a}t der Ans{\"a}tze zur Modellierung und Verifikation werden in verschiedenen Fallstudien aus den Bereichen der reaktiven Systeme und der Kommunikationsprotokolle demonstriert.}, subject = {Verteiltes System; Verifikation}, language = {de} } @phdthesis{Simon2001, author = {Simon, Frank}, title = {Meßwertbasierte Qualit{\"a}tssicherung - ein generisches Distanzmaß zur Erweiterung bisheriger Softwareproduktmaße}, url = {http://nbn-resolving.de/urn:nbn:de:kobv:co1-000000124}, school = {BTU Cottbus - Senftenberg}, year = {2001}, abstract = {In dieser Arbeit werden neue Ans{\"a}tze zur meßwertbasierten Qualit{\"a}tssicherung von Softwareprodukten f{\"u}r praxisrelevante Systemgr{\"o}ßen beschrieben. Dies umfaßt Hilfen zur spezifischen Definition von Qualit{\"a}t, Prozesse zur Qualit{\"a}tssicherung, Werkzeugunterst{\"u}tzung sowie die Identifikation von Problemen bei der Anwendung vieler bisheriger Softwaremaße. Die pragmatische Anwendung der Meßtheorie erlaubt eine Erweiterung in Form eines auf {\"A}hnlichkeiten zwischen Softwareelementen basierenden generischen Distanzmaßes. Dessen unterschiedliche Instanziierungen erm{\"o}glichen die automatische, anpaßbare Gruppierung von Elementen, sowie die ausdrucksstarke, ausschließlich durch Meßwerte bestimmte Softwarevisualisierung. Die ebenfalls in dieser Arbeit wiedergegebene empirische Arbeit, in der die Effizienz und Effektivit{\"a}t dieses Ansatzes mehrfach belegt wird, demonstriert die m{\"o}gliche Bedeutung der meßwertbasierten Qualit{\"a}tssicherung bei der Analyse großer Systeme.}, subject = {Softwareentwicklung; Qualit{\"a}tssicherung; Qualit{\"a}tssteigerung; Softwaremetrie; Softwarequalit{\"a}t; Softwarevermessung; Softwaremaße; Softwarewartung}, language = {de} } @phdthesis{Steinbrueckner2013, author = {Steinbr{\"u}ckner, Frank}, title = {Consistent Software Cities : supporting comprehension of evolving software systems}, url = {http://nbn-resolving.de/urn:nbn:de:kobv:co1-opus-28078}, school = {BTU Cottbus - Senftenberg}, year = {2013}, abstract = {Software visualization is a field of software engineering which aims at providing visual representations of software systems or particular aspects thereof. Numerous approaches for the visualization of software systems have been developed during the last decades. Software cities denote one particular kind of software visualizations that adopts the city metaphor for depicting software systems as virtual cities. Due to their high expressiveness and effectiveness software cities are mainly used for program comprehension tasks, during reverse engineering, and in quality analysis scenarios. Software systems evolve. They are steadily enhanced to provide new product functions, corrected to fix defects, or adapted to changed system environments. Software cities, as proposed today, do not take this evolutionary character of software systems into account. The evolution of software systems, i.e. changes to their internal structure and other characteristics, can easily disrupt the overall software city structure and thus yield very inconsistent visualizations for evolving software systems. The interpretation of these evolving software cities may be error-prone and time-consuming. In this thesis we propose a new software city approach that is based on the observation that real cities often reveal their evolution insofar as they contain e.g. historic centers or satellite cities. Analogous patterns, however, cannot be found in software cities which is an astonishing divergence. We claim that by preserving historical structures and carefully expanding software cities during software evolution, i.e. by spatializing software evolution, we obtain highly expressive, effective, and consistent software cities which support a broader range of application scenarios than the current state of the field. The main contribution of this thesis is a new layout approach for software cities, which explicitly takes evolution into account. Its effects are twofold: First, evolution becomes directly visible in the software city structure in the form of specific geographic patterns that each depicts particular evolutionary phenomena. The resulting increased expressiveness allows for supporting new application scenarios which are evaluated for several example systems. Second, software evolution does no longer disrupt the overall software city structure. Instead, the software cities evolve smoothly during system evolution, which allows for using them during ongoing system development and maintenance. The high consistency of this approach is confirmed in an empirical evaluation.}, subject = {Informationstechnik; Stadtentwicklung; Virtuelle Realit{\"a}t; Rechnernetz; Softwarest{\"a}dte; Softwarevisualisierung; Software cities; Software visualization}, language = {en} } @phdthesis{Noack2007, author = {Noack, Andreas}, title = {Unified quality measures for clusterings, layouts, and orderings of graphs, and their application as software design criteria}, url = {http://nbn-resolving.de/urn:nbn:de:kobv:co1-opus-4046}, school = {BTU Cottbus - Senftenberg}, year = {2007}, abstract = {How good is a given graph clustering, graph layout, or graph ordering --specifically, how well does it group densely connected vertices and separate sparsely connected vertices? How good is a given software design -- specifically, how well does it minimize the interdependence of the subsystems? This work introduces and validates simple and uniform measures for these two properties. Together with existing optimization algorithms, the introduced measures enable the automatic computation e.g. of communities in social networks and of design flaws in software systems. The first part derives, validates, and unifies quality measures for graph clusterings, graph layouts, and graph orderings, with the following results: - Identical quality measures can be applied to clusterings, layouts, and orderings; this enables the computation of consistent clusterings, layouts, and orderings. - Diverse existing and new measures can be unified into few general measures; this facilitates their comparison and validation. - Many existing measures are biased towards certain clusterings, layouts, or orderings, even for graphs without particularly dense or sparse subgraphs, and thus do not (only) measure quality in the above sense. - For example graphs, the minimization of new, unbiased (or weakly biased) measures reveals nonobvious groups, e.g. communities in social networks, subject areas in hypertexts, or closely interlocked countries in international trade. The second part derives, validates, and unifies dependency-based indicators of software design quality. It applies two quality measures for graph clusterings as measures for the coupling of software subsystems -- specifically for the coupling indicated by common changes and for the coupling indicated by references -- and shows: - The measures quantify the dependency-caused development costs, under well-defined simplifying assumptions. - The minimization of the measures conforms to existing dependency-related design principles (like locality of change, acyclicity of references, and stability of references), design rules, and design patterns. - For example software systems, the incremental minimization of the measures reveals nonobvious design flaws, like the distribution of coherent responsibilities over several subsystems, or references from low-level to high-level subsystems. In summary, this work shows that - simple measures can suffice to capture important aspects of graph clustering quality, graph layout quality, graph ordering quality, and software design quality, and - the optimization of simple measures can suffice to detect nonobvious and often useful structure in various real-world systems.}, subject = {Softwareentwicklung; Graphische Darstellung; Computergraphik; Projektmanagement; Graph-Clustering; Graphenzeichnen; Wissensextraktion; Softwareentwurf; Refactoring; Graph clustering; Graph layout; Knowledge discovery; Software design; Refactoring}, language = {en} }