191 research outputs found
On the equilibrium problem and infinitesimal mechanisms of class theta tensegrity systems
No full textThis work presents a study on the equilibrium problem and the infinitesimal mechanisms of class θ= 1 tensegrity prisms. Local solutions of the self-equilibrium problem are numerically obtained through Newton-Raphson iterations. The presented results suggest that the analyzed structures can be usefully employed as building blocks of novel tensegrity metamaterials, due to their rich kinematic response and the considerably large number of infinitesimal mechanisms. © 2019 Author(s)
Supplemental Material, 20180718ZaganasetalCretanAgingCohortSupplementaryRevised - The Cretan Aging Cohort: Cohort Description and Burden of Dementia and Mild Cognitive Impairment
No full textSupplemental Material, 20180718ZaganasetalCretanAgingCohortSupplementaryRevised for The Cretan Aging Cohort: Cohort Description and Burden of Dementia and Mild Cognitive Impairment by Ioannis V. Zaganas, Panagiotis Simos, Maria Basta, Stefania Kapetanaki, Symeon Panagiotakis, Irini Koutentaki, Nikolaos Fountoulakis, Antonios Bertsias, George Duijker, Chariklia Tziraki, Nikolaos Scarmeas, Andreas Plaitakis, Dimitrios Boumpas, Christos Lionis, and Alexandros N. Vgontzas in American Journal of Alzheimer's Disease & Other Dementias</p
Goodness-of-fit tests in conditional duration models
No full textWe propose specification tests for the innovation distribution in conditional duration models. The new tests are based either on the cumulative distribution function, or on exponential transforms such as the Laplace transform and the characteristic function, or on characterizations of the innovation-distribution under test. We study the finite-sample performance of the proposed procedures in comparison with alternative tests which employ nonparametric density estimates as well as with tests based on entropy. A bootstrap version of the tests is utilized in order to study the small sample behavior of the procedures. A real-data example illustrates the applicability of our method and confirms conclusions drawn by earlier author
Kombinatorische und Algorithmische Konstruktionen von Covering Arrays
No full textCovering Arrays sind kombinatorische Designs. Als solche werden diese üblicherweise als Matrizen mit speziellen Eigenschaften betreffend des Vorkommens von Tupeln in gewissen Teilmatrizen definiert. Ziel dieser Arbeit ist es, eine Einführung in Covering Arrays und deren Generalisierungen zu geben, um im Anschluss sowohl kombinatorische als auch algorithmische Konstruktionsmethoden dieser Strukturen zu diskutieren. Im Verlauf dieser Diskussion werden verschiedenste Verbindungen zu anderen Teilbereichen der diskreten Mathematik, wie Gruppentheorie und endliche Körper, hergestellt und angewandt. Bei dem Studium von Covering Arrays ergibt sich das zentrale Problem, optimale Covering Arrays, das sind solche mit der geringsten Anzahl an Zeilen, zu erzeugen. Oft muss das Ziel, Covering Arrays mit der geringsten Anzahl an Zeilen zu finden, aufgegeben und durch ein Streben nach solchen mit einer geringen Anzahl an Zeilen ersetzt werden. Dies zeigt der aktuelle Stand der Forschung, nach welchem Konstruktionen für optimale Covering Arrays nur für spezielle Klassen bekannt sind. Das Generieren optimaler Covering Arrays ist nicht nur aus theoretischer Sicht ein interessantes Problem, sondern auch von praktischem Interesse, da Covering Arrays in Testverfahren, vor allem im Bereich automa-tischer Softwaretests, Anwendung finden.Covering arrays are discrete structures appearing in combinatorial design theory. Most frequently, they are introduced as arrays having specific coverage properties regarding the appearance of tuples in certain subarrays. The aim of this thesis is not only to give a thorough introduction to covering arrays and some of their generalizations, but also to describe combinatorial and algorithmic constructions of these structures. In doing so, links to various fields of discrete mathematics such as group theory and the theory of finite fields are established. Throughout the whole thesis, the reader will be guided by an objective for optimality, as one notorious problem that arises is to find covering arrays that have the smallest number of rows. Often the concept of optimality has to be replaced by the aim for covering arrays that have a small number of rows, as the current state of the art is that constructions of optimal covering arrays are only known for some special classes of covering arrays. The generation of optimal covering arrays is not only a theoretically interesting problem, but is also of interest for practical purposes, as covering arrays find applications in testing, especially in automated software testing
Design theory methods and their applications to the science of security
No full textThis Thesis introduces the research program Design Theory Framework for the Science of Security (DEFSYS) and presents its application to four areas of information security. The primary line of presented DEFSYS-activities concerns vulnerability research problems in software security, approached from a software (security) testing point of view, targeting the three areas of system security, injection attacks and security protocol testing. The secondary line of presented DEFSYS-activities evolves around the area of online privacy research, specifically targeting browser fingerprinting. In all applications of DEFSYS given this Thesis, we concentrate on pointing out how discrete mathematical models can be used synergistically together with combinatorial methods to address problems in information security. Two inherent properties of the employed combinatorial design structures play a significant role and provide benefits in all four presented applications within information security: guaranteed coverage (e.g., of tuples or sub-permutations) and – simultaneously – efficiency (e.g., given by reduced test set sizes). Our research on DEFSYS presented in this Thesis has been empirically supported and validated by our achieved combinatorial security testing results in all four given application areas of information security
Combinatorial design theory and applications for software testing
No full textThis thesis pertains to the interplay of combinatorial design theory and software testing. On the one hand, the problem of generating test sets for black-box testing of software systems which are modeled via vectorial input can be abstracted and treated as subject matter of combinatorial design theory and, in a wider sense, as part of discrete mathematics and theoretical computer science. On the other hand, it is subject to this thesis to apply combinatorial designs to solve problems occurring in the realm of software testing. The aforementioned interplay manifests currently in Combinatorial Testing, a software testing methodology centered around test sets that achieve full t-way coverage of a software’s input space. In order to further and strengthen the interconnections between combinatorial design theory for software testing, we will contribute to and advance individual parts of the combinatorial testing process. According to the overall methodology, this thesis is structured in two parts. The first part pertains to the theoretic aspects of this work, such as objects of combinatorial designs, their properties, generation and related problems; the second part comprises concrete applications of combinatorial design theory for testing software and software-aided systems, including dedicated case studies and industrial applications. In the conclusion we capture the individual contributions described throughout this work and outline how they advance the combinatorial testing process, thereby extending the interplay of combinatorial design theory and software testing
Sicherheitstests für das Bluetooth Low Energy Protokoll mit Kombinatorischen Methoden
No full textDer Bluetooth Low Energy (BLE) Standard ist eine weit verbreitete drahtlose Kommunikationstechnologie für Geräte des Internets der Dinge (IoT) und ermöglicht energiesparende Datenübertragung für verschiedene Anwendungen. Da die Verbreitung von BLE rasant zunimmt, wird die Gewährleistung der Sicherheit zum Schutz vor potenziellen Schwachstellen immer wichtiger. Anbieter BLE fähiger Mikrocontroller müssen BLE Protokolle in ihren Geräten entsprechend der Bluetooth Core Specification implementieren. Trotz der Standardisierungsbemühungen wurden in den BLE Implementierungen verschiedener Anbieter mithilfe manueller und automatisierter Methoden Schwachstellen entdeckt, die potentiell Millionen von Geräten betreffen. Dies ist teilweise auf die Komplexität der Protokolle zurückzuführen, die sich aus der überwältigenden Anzahl möglicher Konfigurationen ergibt und darauf, dass ein gründliches Testen der Implementierung aufgrund der Host-Controller-Schnittstelle (HCI) schwierig ist. In den letzten Jahren wurde das GreyHound Fuzzing Framework entwickelt, das mithilfe kostengünstiger Hardware beliebige BLE Pakete bis zur Linklayer Schicht senden kann. Da Fuzzing von Natur aus probabilistisch ist, ersetzen wir die oben genannte Fuzzing Methode durch einen kombinatorischen Sicherheits Test (CST) Ansatz, der eine garantierte Abdeckung des modellierten Eingabebereichs bietet. Durch die Generierung von Testfällen, die mehrere Kombinationen von Eingabeparametern abdecken, wollen wir Schwachstellen identifizieren, die mit herkömmlichen Testmethoden möglicherweise nicht entdeckt werden. Wir evaluieren unseren Ansatz anhand von Tests mit zehn verschiedenen BLE-Geräten mit unterschiedlichen Firmware Versionen. Insgesamt identifizieren wir 19 verschiedene Probleme, reproduzieren Ergebnisse früherer Arbeiten und decken zusätzliche Fehler auf. Um die Wirksamkeit unserer Methode zu überprüfen, vergleichen wir zusätzlich die Leistung unseres CST-Tools mit der des ursprünglichen Fuzzers und vergleichen deren Ausführungszeit und Fehlererkennungsfähigkeiten.The Bluetooth Low Energy (BLE) standard is a widely used wireless communication technology for Internet of Things (IoT) devices, enabling low-power data transmission for various applications. As the adoption of BLE continues to grow rapidly, ensuring their security becomes more and more important to protect against potential vulnerabilities. Vendors of BLE capable micro controllers are required to implement BLE protocols in their manufactured devices compliant to the Bluetooth Core Specification. Despite the efforts of standardization, several vulnerabilities were discovered in the BLE protocol implementations of multiple vendors using manual and automated methods, potentially affecting millions of devices. This can partially be attributed to the protocol's complexity, stemming from an overwhelming number of possible configurations and the fact that it is difficult to test implementations thoroughly due to the Host Controller Interface (HCI). In recent years, the GreyHound fuzzing framework was developed, which is able to send arbitrary BLE packets down to the link layer, using inexpensive consumer hardware. Since fuzzing is inherently probabilistic, we replace the aforementioned fuzzing method with a Combinatorial Security Testing (CST) approach that provides a guaranteed degree of input space coverage over the parameter model. By generating test cases that cover multiple combinations of input parameters, we aim to identify vulnerabilities that may not be uncovered through traditional testing methods. We evaluate our approach by testing 10 different BLE devices with a variety of firmware versions. In total we identify 19 distinct issues, replicating findings of the previous work and uncovering additional faults. To examine the effectiveness of our method, we additionally provide a performance comparison of our CST tool against the original fuzzer, contrasting their execution time and fault detection capabilities
Algebraic methods for experimental design theory
No full textDie Integration von algebraischen Methoden in die Statistik in den frühen und mittleren 1990er Jahren [33, 105] hat beide Forschungsgebiete von den entstandenen Synergien profitieren lassen [4, 104]. Diese Arbeit beschäftigt sich mit kombinatorischen Designs, welche im relativ neuen Bereich des kombinatorischen Testen (KT) für Software als Teilgebiet der statistischen Versuchsplanung verwendet werden [81]. Der Begriff der “Abdeckung”, der als eine Verallgemeinerung des bekannten -fachen Auftretens von t-Tupeln in orthogonalen Arrays angesehen werden kann, steht an zentraler Stelle in dem KT und findet sich auch in den definierenden Eigenschaften der in diesem Bereich betrachteten Strukturen wieder. Zu den betrachteten Strukturen zählen abdeckende Arrays, welche man als spezielle Klasse von kombinatorischen Designs ansehen kann, sowie auch gewisse Klassen von endlichen Sequenzen [28]. Das Ziel dieser Arbeit ist zu analysieren und darstellen, wie algebraische Methoden in der Spezifikation, Erzeugung und der Charakterisierung von Eigenschaften dieser Strukturen verwendet werden können [40]. Die zugrundeliegenden algebraischen Methoden basieren auf Polynomen [16].Since the introduction of algebraic techniques into the field of statistics in the start and middle of the 1990s [33, 105], both fields have immensely benefited from the resulting synergies [4, 104]. This Thesis is concerned with classes of combinatorial designs, that appear in a relatively new subfield called Combinatorial Testing (CT) for Software of Design of Experiments [81]. The notion of “coverage requirement”, which represents a generalization of the well established notion of exactly -way appearance of t-tuples in orthogonal arrays, is fundamental to the field of CT and is also a fundamental property in the discrete structures that are considered in CT. These structures include covering arrays, which can be regarded as a special class of combinatorial designs, and certain classes of finite sequences [28]. The aim of this Thesis is to analyse and depict how algebraic techniques can help in the specification, generation and property assessment of these structures [40]. Polynomial algebraic techniques are the basic methodologies which are to be applied in this domain [16]
Exakte Methoden zur Generierung von Covering Arrays
No full textCovering Arrays sind interessante kombinatorische Objekte, die Bekanntheit erlangt haben durch ihre Anwendbarkeit für kombinatorisches Testen, einem Zweig von Software Testen. Die Generierung von Covering Arrays wird normalerweise als Optimierungsproblem betrachtet, wo es gewünscht ist, ein Covering Array mit einer minimalen Anzahl von Zeilen zu finden. Durch Fortschritte in den vergangenen Jahren wurden exakte Methoden sehr effizient und anpassungsfähig, das heißt sie können bei zahlreichen Problemen angewandt werden. Für Covering Array Generierung mittels exakter Methoden existieren mehrere Kodierungen. Die Anwendung von exakten Methoden für einzelne Teile eines Covering Array Generierungsalgorithmus dagegen wurde nur selten untersucht. Der Zweck dieser Arbeit ist eine Untersuchung der Möglichkeit, Algorithmen zur Erzeugung von Covering Arrays mit exakten Methoden zu verbessern. Im Zuge dieser Arbeit wurden zwei neue Algorithmen entwickelt, die exakte Methoden, genauer gesagt SAT solving, pseudo-Boolean constraint solving und MaxSAT solving, anwenden, um Teilprobleme von Covering Array Generierungsalgorithmen zu lösen. Der vorgestellte ClassifyBalancedCAs Algorithmus kann Covering Arrays klassifizieren, also alle nicht-äquivalenten Covering Arrays einer gegebenen Größe zählen, und diese Menge von Covering Arrays generieren. In einigen Fällen ist der ClassifyBalancedCAs Algorithmus schneller als alle existierenden Algorithmen zur Klassifizierung von Covering Arrays, besonders wenn eine Optimierung namens balancebasiertes Pruning verwendet wird. Zusätzlich wurde der IPO-MAXSAT Algorithmus entwickelt, bei dem MaxSAT verwendet wird, um optimale Lösungen für Teilprobleme der In-Parameter-Order (IPO) Strategie zu finden. Dadurch, dass optimale Lösungen für Teilprobleme verwendet werden, können die Fähigkeiten und Limitierungen der IPO Strategie untersucht werden. Experimente zeigen, dass die Verwendung von optimalen Lösungen für Teilprobleme die Qualität der produzierten Covering Arrays erhöht. Sie zeigen aber auch, dass die generierten Covering Arrays dennoch nicht optimal sind. Die präsentierten Algorithmen demonstrieren, dass es vorteilhaft sein kann, existierende Covering Array Algorithmen mit exakten Methoden zu erweitern.Covering arrays are interesting combinatorial objects that gained much popularity due to their application in combinatorial testing, which is a branch of software testing. The generation of covering arrays is usually treated as an optimization problem, where it is desired to find a covering array with a minimal number of rows. Over the past years exact methods have evolved to be efficient and highly adaptive, meaning they can be applied to many different problems. While several encodings for covering array generation with some exact methods exist, the application of exact methods for only part of a covering array generation algorithm has rarely been studied. The purpose of this thesis is examining the possibility of enhancing covering array generation algorithms with exact methods. In the course of this thesis two new algorithms were developed applying exact methods, in particular SAT solving, pseudo-Boolean constraint solving and MaxSAT solving, to subproblems occurring in algorithms for covering array generation. The proposed algorithm ClassifyBalancedCAs is capable of covering array classification, that is counting all non-equivalent covering arrays of a given size, and exhaustive generation of all such covering arrays. In several cases the ClassifyBalancedCAs algorithm is faster than all existing covering array classification algorithms, especially when an optimization called balance-based pruning is used. Additionally, the IPO-MAXSAT algorithm was developed, where MaxSAT is used to find optimal solutions for subproblems occurring in the greedy In-Parameter-Order (IPO) strategy. Using optimal solutions for subproblems allows to investigate the capabilities and limitations of the IPO strategy. Experiments show that better solutions for subproblems lead to higher quality of the generated covering arrays, however, using optimal solutions for subproblems is not sufficient to generate optimal covering arrays. The presented algorithms show that enhancing covering array generation algorithms with exact methods can be beneficial
Efficient algorithms and tools for practical combinatorial testing
No full textCombinatorial Testing (CT) is an effective testing strategy which has seen increased adoption in the field of software testing. As a black-box testing technique it concerns itself with generating input test cases able to provide mathematical input-space coverage guarantees while keeping the number of test cases low. Generating optimal combinatorial test sets is believed to be a hard optimization problem which has spurred the development of several greedy algorithms which trade-off optimality for speed. One of the most popular representatives is the In-Parameter-Order family of algorithms. This class of algorithms has seen a wide adoption in including software, hardware and security testing as they are able to construct test sets for most instances occurring in practice. In this thesis, an efficient design of the In-Parameter-Order algorithms is presented which drastically reduces the time and required memory to compute combinatorial test sets. The improved theoretical design is implemented in a prototype and benchmarked against the state of the art implementation. Furthermore, the algorithms are made available in a general purpose test generation tool developed for this thesis which is open and free to use by anyone. The thesis also studies the impact of tie-breaking, parameter ordering and tuple enumeration order on the resulting test sets. An extensive experimental study shows that different tie-breakers have no significant impact on average and that ordering parameters in order of descending domain size can lead to drastically smaller test sets
- …
