004 Informatik
Refine
Year of publication
Document Type
- Bachelor Thesis (49)
- Master's Thesis (33)
- Conference Proceeding (6)
- Contribution to a Periodical (5)
- Article (reviewed) (2)
- Doctoral Thesis (2)
- Book (1)
Conference Type
- Konferenzartikel (6)
Keywords
- IT-Sicherheit (12)
- JavaScript (9)
- Computersicherheit (4)
- Deep learning (4)
- E-Learning (4)
- HTML 5.0 (4)
- Internet der Dinge (4)
- Internet of Things (4)
- Blockchain (3)
- Künstliche Intelligenz (3)
Institute
- Fakultät Medien (M) (ab 22.04.2021) (36)
- Fakultät Medien und Informationswesen (M+I) (bis 21.04.2021) (36)
- Fakultät Elektrotechnik, Medizintechnik und Informatik (EMI) (ab 04/2019) (17)
- Fakultät Elektrotechnik und Informationstechnik (E+I) (bis 03/2019) (6)
- ivESK - Institut für verlässliche Embedded Systems und Kommunikationselektronik (4)
- Zentrale Einrichtungen (3)
- Fakultät Wirtschaft (W) (2)
- IMLA - Institute for Machine Learning and Analytics (1)
Open Access
- Closed Access (47)
- Open Access (27)
- Closed (24)
- Diamond (2)
Conceptualization and implementation of automated optimization methods for private 5G networks
(2023)
Today’s companies are adjusting to the new connectivity realities. New applications require more bandwidth, lower latency, and higher reliability as industries become more distributed and autonomous. Private 5th Generation (5G) networks known as 5G Non-Public Networks (5G-NPN), is a novel 3rd Generation Partnership Project (3GPP)- based 5G network that can deliver seamless and dedicated wireless access for a particular industrial use case by providing the mentioned application’s requirements. To meet these requirements, several radio-related aspects and network parameters should be considered. In many cases, the behavior of the link connection may vary based on wireless conditions, available network resources, and User Equipment (UE) requirements. Furthermore, Optimizing these networks can be a complex task due to the large number of network parameters and KPIs that need to be considered. For these reasons, traditional solutions and static network configuration are not affordable or simply impossible. Despite the existence of papers in the literature that address several optimization methods for cellular networks in industrial scenarios, more insight into these existing but complex or unknown methods is needed.
In this thesis, a series of optimization methods were implemented to deliver an optimal configuration solution for a 5G private network. To facilitate this implementation, a testing system was implemented. This system enables remote control over the UE and 5G network, establishment of a test environment, extraction of relevant KPI reports from both UE and network sides, assessment of test results and KPIs, and effective utilization of the optimization and sampling techniques.
The research highlights the advantageous aspects of automated testing by using OFAT, Simulated Annealing, and Random Forest Regressor methods. With OFAT, as a common sampling method, a sensitivity analysis and an impact of each single parameter variation on the performance of the network were revealed. With Simulated Annealing, an optimal solution with MSE of roughly 10 was revealed. And, in the Random Forest Regressor, it was seen that this method presented a significant advantage over the simulated annealing method by providing substantial benefits in time efficiency due to its machine- learning capability. Additionally, it was seen that by providing a larger dataset or using some other machine-learning techniques, the solution might be more accurate.
Künstliche Intelligenzen, Deep Learning und Machine-Learning-Algorithmen sind im digitalen Zeitalter zu einem Punkt gekommen, in dem es schwer ist zu unterscheiden, welche Informationen und Quellen echt sind und welche nicht. Der Begriff „Deepfakes“ wurde erstmals 2017 genutzt und hat bereits 2018 mit einer App bewiesen, wie einfach es ist, diese Technologie zu verwenden um mit Videos, Bildern oder Ton Desinformationen zu verbreiten, politische Staatsoberhäupter nachzuahmen oder unschuldige Personen zu deformieren. In der Zwischenzeit haben sich Deepfakes bedeutend weiterentwickelt und stellen somit eine große Gefahr dar.
Diese Arbeit bietet eine Einführung in das Themengebiet Deepfakes. Zudem behandelt sie die Erstellung, Verwendung und Erkennung von Deepfakes, sowie mögliche Abwehrmaßnahmen und Auswirkungen, welche Deepfakes mit sich bringen.
As e-commerce platforms have grown in popularity, new difficulties have emerged, such as the growing use of bots—automated programs—to engage with e-commerce websites. Even though some algorithms are helpful, others are malicious and can seriously hurt e-commerce platforms by making fictitious purchases, posting fictitious evaluations, and gaining control of user accounts. Therefore, the development of more effective and precise bot identification systems is urgently needed to stop such actions. This thesis proposes a methodology for detecting bots in E-commerce using machine learning algorithms such as K-nearest neighbors, Decision Tree, Random Forest, Support Vector Machine, and Neural Network. The purpose of the research is to assess and contrast the output of these machine learning methods. The suggested approach will be based on data that is readily accessible to the public, and the study’s focus will be on the research of bots in e-commerce.
The purpose of the study is to provide an overview of bots in e-commerce, as well as information on the different kinds and traits of bots, as well as current research on bots in e-commerce and associated work on bot detection in e-commerce. The research also seeks to create a more precise and effective bot detection system as well as find critical factors in detecting bots in e-commerce.
This research is significant because it sheds light on the increasing issue of bots in e-commerce and the requirement for more effective bot detection systems. The suggested approach for using machine learning algorithms to identify bots in ecommerce can give e-commerce platforms a more precise and effective bot detection system to stop malicious bot activities. The study’s results can also be used to create a more effective bot detection system and pinpoint key elements in detecting bots in e-commerce.
In dieser Arbeit wird der Bildbearbeitungsprozess von Dokumenten mithilfe von einem schlicht gehaltenem Neuronalen Netzwerk und Bearbeitungsoperationen optimiert. Ziel ist es, abfotografierte Dokumente zum Drucken aufzubereiten, sodass die Schrift gut lesbar, gerade und nicht verzerrt ist und Störfaktoren herausgefiltert werden. Als API zur Verfügung gestellt, können Bilder von Dokumenten beliebiger Größe und Schriftgröße bearbeitet werden. Während ein unter schlechten Bedingungen schräg aufgenommenes Bild nach Tesseract keine Buchstaben enthält, wird mit dem bearbeiteten Bild davon eine Buchstabenfehlerrate von 0,9% erreicht.
Die Komplexität von Softwareprojekten hat in den letzten Jahren stetig zugenommen. Um den gleichzeitig steigenden Anforderungen an die Codequalität gerecht zu werden, setzen auch ursprünglich dynamisch typisierte Programmiersprachen zuhnemend auf statische Typisierung. Dies kann in Form von externen Werkzeugen geschehen, die zusätzlich zum eigentlichen Compiler den Code auf Typsicherheit überprüfen, oder alternativ durch Erweiterung der Compiler selbst, um die Unterstützung für statische Typisierung direkt in der Sprache zu verankern. Ziel des etylizer-Projekts ist es, für die Programmiersprache Erlang zunächst ein solches externes Tool bereitzustellen und langfristig Teil der Compiler-Toolchain zu werden.In dieser Arbeit wird der Typchecker um die Fähigkeit erweitert, Erlang-Projekte vollständig zu verifizieren. Dafür wird zunächst die interne Symboltabelle erweitert, die etylizer nutzt, um Verweise auf Funktionen und Typen aus anderen Modulen aufzulösen. Die Implementierung der Symboltabelle wird so angepasst,dass sie zur Laufzeit um alle für das aktuell geprüfte Modul benötigten Symbole erweitert wird. Um die Laufzeit im Rahmen zu halten, wird ein Algorithmus entwickelt, der die Abhängigkeiten zwischen den Source-Code Dateien des Erlang-Projekts erkennt und anhand dieser entscheidet, welche Dateien sich seit dem letzten Durchlauf geändert haben und deshalb erneut überprüft werden müssen.
The Internet of Things is spreading significantly in every sector, including the household, a variety of industries, healthcare, and emergency services, with the goal of assisting all of those infrastructures by providing intelligent means of service delivery. An Internet of Vulnerabilities (IoV) has emerged as a result of the pervasiveness of the Internet of Things (IoT), which has led to a rise in the use of applications and devices connected to the IoT in our day-to-day lives. The manufacture of IoT devices are growing at a rapid pace, but security and privacy concerns are not being taken into consideration. These intelligent Internet of Things devices are especially vulnerable to a variety of attacks, both on the hardware and software levels, which leaves them exposed to the possibility of use cases. This master’s thesis provides a comprehensive overview of the Internet of Things (IoT) with regard to security and privacy in the area of applications, security architecture frameworks, a taxonomy of various cyberattacks based on various architecture models, such as three-layer, four-layer, and five-layer. The fundamental purpose of this thesis is to provide recommendations for alternate mitigation strategies and corrective actions by using a holistic rather than a layer-by-layer approach. We discussed the most effective solutions to the problems of privacy and safety that are associated with the Internet of Things (IoT) and presented them in the form of research questions. In addition to that, we investigated a number of further possible directions for the development of this research.
As cyber threats continue to evolve, it is becoming increasingly important for organizations to have a Security Operations Center (SOC) in place to effectively defend against them. However, building and maintaining a SOC can be a daunting task without clear guidelines, policies, and procedures in place. Additionally, most current SOC solutions used by organizations are outdated, lack key features and integrations, and are expensive to maintain and upgrade. Moreover, proprietary solutions can lead to vendor lock-in, making it difficult to switch to a different solution in the future.
To address these challenges, this thesis proposes a comprehensive SOC framework and an open-source SOC solution that provides organizations with a flexible and cost-effective way to defend against modern cyber threats. The research methodology involved conducting a thorough literature review of existing literature and research on building and maintaining a SOC, including using SOC as a service. The data collected from the literature review was analyzed to identify common themes, challenges, and best practices for building and maintaining a SOC.
Based on the data collected, a comprehensive framework for building and maintaining a SOC was developed. The framework addresses essential areas such as the scope and purpose of the SOC, governance and leadership, staffing and skills, technologies and tools, processes and procedures, service level agreements (SLAs), and evaluation and measurement. This framework provides organizations with the necessary guidance and resources to establish and effectively operate a SOC, as well as a reference for evaluating the service provided by SOC service providers.
In addition to the SOC framework, a modern open-source SOC solution was developed, which emphasizes several key measures to help organizations defend against modern cyber threats. These measures include real-time, actionable threat intelligence, rapid and effective incident response, continuous security monitoring and alerting, automation, integration, and customization. The use of open-source technologies and a modular architecture makes the solution cost-effective, allowing organizations to scale it up or down as needed.
Overall, the proposed SOC framework and open-source SOC solution provide organizations with a comprehensive and systematic approach for building and maintaining a SOC that is aligned with the needs and objectives of the organization. The open-source SOC solution provides a flexible and cost-effective way to defend against modern cyber threats, helping organizations to effectively operate their SOC and reduce their risk of security incidents and breaches.
In dieser Forschungsarbeit wird die Datensicherheit von Microsoft Azure analysiert und bewertet. Die Bewertung findet dabei aus der Sicht von Unternehmen statt. Im ersten Abschnitt wird zunächst der grundlegende Aufbau und die unterschiedlichen Formen des Cloud Computing beschrieben. Im zweiten Teil wird ein Vergleich der drei größten Cloud Anbieter vollzogen. Der letzte Teil besteht aus der Evaluation der Datensicherheit von Azure, wobei auf Aspekte wie Datenschutz, Bedrohungen und Schutzmaßnahmen eingegangen wird. Abschließend wird eine Empfehlung für das Unternehmen Bechtle GmbH Offenburg IT-Systemhaus abgegeben.
Im Verlauf der Arbeit stellt sich heraus, dass Azure eine ausreichende Datensicherheit bieten kann. Allerdings wird deutlich, dass durch die Kombination von mehreren Nebenfaktoren wie das Patch-Verhalten oder die Antwortzeit auf Sicherheitsschwachstellen seitens Microsofts, eine große Gefahr für die Daten von Unternehmen entstehen kann. Demnach ist Microsoft als Anbieter ein größeres Problem für die Sicherheit von Daten in Azure als der Cloud-Dienst selbst.
Die Thesis beschäftigt sich mit dem Kommunikationsprotokoll Lightweight Machine to Machine, welches für das Internet of Things entwickelt wurde. Es soll untersucht werden, wie das Protokoll funktioniert und wie es eingesetzt werden kann. Ebenfalls soll die Thesis zeigen, wie und ob Lightweight Machine to Machine über Long Term Evolution for Machines für Anwendungen mit begrenzten Ressourcen geeignet ist. Um diese Fragestellung zu beantworten, wurde das Protokoll auf Grund seiner Spezifikation und seinen Softwareimplementationen untersucht. Daraufhin wurde ein Versuchssystem entworfen und dieses anschließend auf sein Laufzeitverhalten und auf sein Energieverbrauch getestet. Die Evaluation des Protokolls ergab, dass es viele sinnvolle Funktionen zugeschnitten auf Geräte im Internet of Things besitzt und diese Funktionen kompakt und verständlich umsetzt. Da das Protokoll noch relativ jung ist, stellt es an verschiedenen Punkten eine Herausforderung dar. Die Tests des Versuchssystems ergaben, dass Lightweight Machine to Machine sich unter bestimmten Bedingungen für ressourcenbegrenzte Anwendungen eignet.
Licht war für die Menschheit schon immer ein Hilfsmittel zur Orientierung. Das Zusammenspiel zwischen hellen und schattierten Oberflächen macht eine räumliche Wahrnehmung erst möglich. Die Lokalisierung von Lichtquellen bietet darüber hinaus für zahlreiche Anwendungsfelder, wie beispielsweise Augmented Reality, ein großes Potential.
Das Ziel der vorliegenden Arbeit war es, ein neuronales Netzwerk zu entwickeln, welches mit Hilfe eines selbst generierten, synthetischen Datensatzes eine Lichtsetzung parametrisiert. Dafür wurden State-of-the-Art Netzwerke aus der digitalen Bildverarbeitung eingesetzt.
Zu Beginn der Arbeit mussten die Eigenschaften der Lichtsetzung extrahiert werden. Eine weitere fundamentale Anforderung war die Aufbereitung des Wissens von Deep Learning.
Für die Generierung des synthetischen Datensatzes wurde eigens ein Framework entwickelt, welches auf der Blender Engine basiert.
Anschließend wurden die generierten Bilder und Metadaten in einem abgewandelten VGG16- und ResNet50-Netz trainiert, validiert und evaluiert.
Eine gewonnene Erkenntnis ist, dass sich künstlich generierte Daten eignen um ein neuronales Netz zu trainieren. Des Weiteren konnte gezeigt werden, dass sich mit Hilfe von Deep Learning Lichtsetzungsparameter extrahieren lassen.
Eine weiterführende Forschungsaufgabe könnte mit dem vorgeschlagenen Ansatzdie Lichtinszenierung von Augmented Reality Anwendungen verbessern.
Organizations striving to achieve success in the long term must have a positive brand image which will have direct implications on the business. In the face of the rising cyber threats and intense competition, maintaining a threat-free domain is an important aspect of preserving that image in today's internet world. Domain names are often near-synonyms for brand names for numerous companies. There are likely thousands of domains that try to impersonate the big companies in a bid to trap unsuspecting users, usually falling prey to attacks such as phishing or watering hole. Because domain names are important for organizations for running their business online, they are also particularly vulnerable to misuse by malicious actors. So, how can you ensure that your domain name is protected while still protecting your brand identity? Brand Monitoring, for example, may assist. The term "Brand Monitoring" applies only to keep tabs on an organization's brand performance, reception, and overall online presence through various online channels and platforms [1]. There has been a rise in the need of maintaining one's domain clear of any linkages to malicious activities as the threat environment has expanded. Since attackers are targeting domain names of organizations and luring unsuspecting users to visit malicious websites, domain monitoring becomes an important aspect. Another important aspect of brand abuse is how attackers leverage brand logos in creating fake and phishing web pages. In this Master Thesis, we try to solve the problem of classification of impersonated domains using rule-based and machine learning algorithms and automation of domain monitoring. We first use a rule-based classifier and Machine Learning algorithms to classify the domains gathered into two buckets – "Parked" and "Non-Parked". In the project's second phase, we will deploy object detection models (Scale Invariant Feature Transform - SIFT and Multi-Template Matching – MTM) to detect brand logos from the domains of interest.
Even though the internet has only been there for a short period, it has grown tremendously. To- day, a significant portion of commerce is conducted entirely online because of increased inter- net users and technological advancements in web construction. Additionally, cyberattacks and threats have expanded significantly, leading to financial losses, privacy breaches, identity theft, a decrease in customers’ confidence in online banking and e-commerce, and a decrease in brand reputation and trust. When an attacker pretends to be a genuine and trustworthy institution, they can steal private and confidential information from a victim. Aside from that, phishing has been an ongoing issue for a long time. Billions of dollars have been shed on the global economy. In recent years, there has been significant progress in the development of phishing detection and identification systems to protect against phishing attacks. Phishing detection technologies frequently produce binary results, i.e., whether a phishing attempt was made or not, with no explanation. On the other hand, phishing identification methodologies identify phishing web- pages by visually comparing webpages with predetermined authentic references and reporting phishing together with its target brand, resulting in findings that are understandable. However, technical difficulties in the field of visual analysis limit the applicability of currently available solutions, preventing them from being both effective (with high accuracy) and efficient (with little runtime overhead). Here, we evaluate existed framework called Phishpedia. This hybrid deep learning system can recognize identity logos from webpage screenshots and match logo variants of the same brand with high precision. Phishpedia provides high accuracy with low run- time. Lastly, unlike other methods, Phishpedia does not require training on any phishing sam- ples whatsoever. Phishpedia exceeds baseline identification techniques (EMD, PhishZoo, and LogoSENSE), inaccurately detecting phishing pages in lengthy testing using accurate phishing data. The effectiveness of Phishpedia was tested and compared against other standard machine learning algorithms and some state-of-the-art algorithms. The given solutions performed better than different algorithms in the given dataset, which is impressive.
Technology advancement has played a vital role in business development; however, it has opened a broad attack surface. Passwords are one of the essential concepts used in applications for authentication. Companies manage many corporate applications, so the employees must meet the password criteria, which leads to password fatigue. This thesis addressed this issue and how we can overcome this problem by theoretically implementing an IAM solution. In this, we disused MFA, SSO, biometrics, strong password policies and access control. We introduced the IAM framework that should be considered while implementing the IAM solution. Implementing an IAM solution adds an extra layer of security.
Eine neue Programmiersprache zu erlernen kann für Anfänger:innen manchmal schwer sein, selbst für Programmiersprachen wie Python, die bekannt dafür sind Einsteigerfreundlich zu sein. Denn selbst wenn die Syntax eines Python Programms schnell verstanden wird, ist oft nicht direkt erkenntlich wie der Code hinter dem Programm funktioniert. Anfänger:innen können dabei auch auf ihre Grenzen stoßen, den Ablauf eines Programmes nur alleine durch den Programmcode zu verstehen. Denn der Text der den Code ausmacht, kann auch nur bis zu einem gewissen Grad vermitteln wie oder was genau abläuft. Um den Ablauf eines Programms besser vermitteln zu können, wird der Code oft z.B. mit Diagrammen visualisiert. Visuelle Elemente können ebenfalls zusätzlich zum Code mehr Unterstützung leisten. Das Thema dieser Arbeit beschäftigt sich mit der Visualisierung von Python Programmen in der Entwicklungsumgebung Visual Studio Code, um Programmieranfänger:innen und Student:innen beim Erlernen der Programmiersprache Python zu unterstützen. Die Entwicklung der Visualisierung beinhaltet, das Erstellen einer Erweiterung in Visual Studio Code, die unter anderem das Debug Adapter Protocol einsetzt um mit dem Python Debugger zu kommunizieren.
Encryption techniques allow storing and transferring of sensitive information securely by using encryption at rest and encryption in transit, respectively. However, when computation is performed on these sensitive data, the data needs to be decrypted first and encrypted again after performing the computations. During the computations, the sensitive data becomes vulnerable to attackers as it's in decrypted form. Homomorphic encryption, a special type of encryption technique that allows computation on encrypted data can be used to solve the above-mentioned problem. The best way to achieve maximum security with homomorphic encryption is to perform at least the homomorphic encryption and decryption on the client side (browser) of a web application by not trusting the server. At present time there are many libraries with different homomorphic schemes available for homomorphic encryption. However, there are very few to no JavaScript libraries available to perform homomorphic encryption on the client side of any web application. This thesis mainly focuses on the JavaScript implementation of client-side homomorphic encryption. The fully homomorphic encryption scheme BFV is selected for the implementation. After implementing the fully homomorphic encryption scheme based on the “py-fhe” library, tests are also carried out in order to determine the applicability (in terms of time consumption, security and correctness) of this implementation in a web application by comparing the performance and security for different test cases and different settings.
Risk-based Cybermaturity Assessment Model - Protecting the company against ransomware attacks
(2023)
Ransomware has become one of the most catastrophic attacks in the previous decade, hurting businesses of all sorts worldwide. So, no organization is safe, and most companies are reviewing their ransomware defensive solutions to avoid business and operational hazards. IT departments are using cybersecurity maturity assessment frameworks like CMMC, C2M2, CMMI, NIST, CIS, CPP, and others to analyze organization security capabilities. In addition to maturity assessment models for the process layer and human pillar, there are much research on the analysis, identification, and defense of cyber threats in product/software layers that propose state-of-the-art approaches.
This motivates a comprehensive ransomware cyber security solution. Then, a crucial question arises: “How companies can measure the security maturity of controls in a specific danger for example for Ransomware attack?” Several studies and frameworks addressed this subject.
Complexity of understanding the ransomware attack, Lack of comprehensive ransomware defense solutions and Lack of cybermaturity assessment model for ransomware defense solutions are different aspects of problem statement in this study. By considering the most important limitations to developing a ransomware defense cybermaturity assessment method, this study developed a cybermaturity assessment methodology and implemented a Toolkit to conduct cyber security self-assessment specifically for ransomware attack to provide a clearer vision for enterprises to analyze the security maturity of controls regardless of industry or size.
Durch die Fortschritte im Bereich der Quantencomputer rückt der Zeitpunkt näher, dass Quantencomputer die bestehenden mathematischen Probleme lösen können, welche in den aktuellen Public-Key-Verschlüsselungsverfahren verwendet werden. Als Reaktion darauf wurde ein Standardisierungsprozess für quantensichere Public-Key-Verschlüsselungsverfahren gestartet. Diese Arbeit analysiert diese und vergleicht sie untereinander, um Stärken und Schwächen der einzelnen Verfahren aufzuzeigen.
On a regular basis, we hear of well-known online services that have been abused or compromised as a result of data theft. Because insecure applications jeopardize users' privacy as well as the reputation of corporations and organizations, they must be effectively secured from the outset of the development process. The limited expertise and experience of involved parties, such as web developers, is frequently cited as a cause of risky programs. Consequently, they rarely have a full picture of the security-related decisions that must be made, nor do they understand how these decisions affect implementation accurately.
The selection of tools and procedures that can best assist a certain situation in order to protect an application against vulnerabilities is a critical decision. Regardless of the level of security that results from adhering to security standards, these factors inadvertently result in web applications that are insufficiently secured. JavaScript is a language that is heavily relied on as a mainstream programming language for web applications with several new JavaScript frameworks being released every year.
JavaScript is used on both the server-side in web applications development and the client-side in web browsers as well.
However, JavaScript web programming is based on a programming style in which the application developer can, and frequently must, automatically integrate various bits of code from third parties. This potent combination has resulted in a situation today where security issues are frequently exploited. These vulnerabilities can compromise an entire server if left unchecked. Even though there are numerous ad hoc security solutions for web browsers, client-side attacks are also popular. The issue is significantly worse on the server side because the security technologies available for server-side JavaScript application frameworks are nearly non-existent.
Consequently, this thesis focuses on the server-side aspect of JavaScript; the development and evaluation of robust server-side security technologies for JavaScript web applications. There is a clear need for robust security technologies and security best practices in server-side JavaScript that allow fine-grained security.
However, more than ever, there is this requirement of reducing the associated risks without hindering the web application in its functionality.
This is the problem that will be tackled in this thesis: the development of secure security practices and robust security technologies for JavaScript web applications, specifically, on the server-side, that offer adequate security guarantees without putting too many constraints on their functionality.
As information technology continues to advance at a rapid speed around the world, new difficulties emerge. The growing number of organizational vulnerabilities is among the most important issues. Finding and mitigating vulnerabilities is critical in order to protect an organization’s environment from multiple attack vectors.
The study investigates and comprehends the complete vulnerability management process from the standpoint of the security officer job role, as well as potential improvements. Few strategies are used to achieve efficient mitigation and the de- velopment of a process for tracking and mitigating vulnerabilities. As a result, a qualitative study is conducted in which the objective is to create a proposed vulner- ability and risk management process, as well as to develop a system for analyzing and tracking vulnerabilities and presenting the vulnerabilities in a graphical dash- board format. This thesis’s data was gathered through an organized literature study as well as through the use of various web resources. We explored numerous ap- proaches to analyze the data, such as categorizing the vulnerabilities every 30, 60, and 90 days to see whether the vulnerabilities were reoccurring or new. According to our findings, tracking vulnerabilities can be advantageous for a security officer.
We come to the conclusion that if an organization has a proper vulnerability tracking system and vulnerability management process, it can aid security officers in having a better understanding of and making plans for reducing vulnerabilities. In terms of system patching and vulnerability remediation, it will also assist the security officer in identifying areas of weakness in the process. As a result, the suggested ways provide an alternate approach to managing and tracking vulnerabilities in an effective manner, although there is still a small area that needs additional analysis and research to make it even better.
Komplexe E-Commerce-Systeme müssen heutzutage immer schneller am Markt sein und sich an diesen anpassen. Dies wird durch SaaS-Services möglich, wodurch sich die Best-of-Breed-Lösungen einsetzen lassen. Der monolithische Ansatz der meisten E-Commerce-Systeme ist für diese Anwendungen nicht mehr geeignet. Abhilfe soll der Composable-Commerce-Ansatz schaffen. Für den Ansatz wird eine Integrationslösung benötigt. Ziel dieser Thesis ist es, Integrationslösungen zu evaluieren und mithilfe von Integration-Layer-Prototypen gegenüberzustellen. Es werden zwei Integrationslösungen ausgewählt, die als Prototyp implementiert werden. Für den ersten Prototypen wird Apache Camel in einem Spring-Boot-Server verwendet. Der zweite Prototyp setzt die AWS-eigenen Services für die Integration ein. Zum Schluss werden diese durch einen Last-Test auf ihre Performance geprüft.
An organized strategy to ensure the security of an organization is an information security management system. During various security crises, hazards, and breaches, this strategy aids an organization in maintaining the confidentiality, integrity, and accessibility of information. Organizations are getting ready to comply with information security management system criteria. Despite this, security concerns continue to plague ineffective controls, have poor connectivity, or cause a silo effect, which is a common cause. One of the causes is a low maturity model that is not synchronized with the organization’s business processes. For a higher level of maturity, it is best to evaluate the practices.
Different maturity models on information security and cyber security capacity, management processes, security controls, implementation level, and many more have already been developed by numerous international organizations, experts, and scholars. The present models, however, do not assess a particular organization's particular practices. The evaluation of the business process is frequently neglected because measurement requirements for models are typically more concentrated on examining specific elements. For this reason, it caused the maturity assessment to not be executed explicitly and broadly.
We developed an organizational information security maturity model, a combination of work of different maturity models currently existing. While making this model, we considered that any size or type of organization could use this model. The model considers the success elements of the information security management system when assessing the implementation's effectiveness. We employed a mixed-method strategy that included both qualitative and quantitative research. With the help of a questionnaire survey, we evaluated the previous research using a qualitative methodology. In the quantitative method, we'll figure out how mature the information security management system is now. The proposed model could be used to reduce security incidents by improving implementation gaps.
Due to the increasing aging of the population, the number of elderly people requiring care is growing in most European countries. However, the number of caregivers working in nursing homes and on daily care services is declining in countries like Germany or Italy. This limits the time for interpersonal communication. Furthermore, as a result of the Covid-19 pandemic, social distancing during contact restrictions became more important, causing an additional reduction of personal interaction. This social isolation can strongly increase emotional stress. Robotic assistance could contribute to addressing this challenge on three levels: (1) supporting caregivers to respond individually to the needs of patients and residents in nursing homes; (2) observing patients’ health and emotional state; (3) complying with high hygiene standards and minimizing human contact if required. To further the research on emotional aspects and the acceptance of robotic assistance in care, we conducted two studies where elderly participants interacted with the social robot Misa. Facial expression and voice analysis were used to identify and measure the emotional state of the participants during the interaction. While interpersonal contact plays a major role in elderly care, the findings reveal that robotic assistance generates added value for both caregivers and patients and that they show emotions while interacting with them.
Software-defined Access (SD-Access, SDA) hat aufgrund der flexiblen, automatisierten und schnelleren Verwaltung von Unternehmensnetzwerken erhebliche Aufmerksamkeit erlangt. Im Gegensatz zu traditionellen Netzwerken mit manuellen Prozessen, bietet SD-Access Zugriffsrichtlinien, Netzwerksegmentierung und Endpunktüberwachung in einer Lösung und trägt damit zur Netzwerksicherheit in Unternehmen bei.
Zunächst wird die SD-Access-Lösung von Cisco vorgestellt und herausgearbeitet, welche Komponenten dafür benötigt werden. Auf dieser Grundlage wird überprüft, welche Voraussetzungen zur Einführung von SD-Access im konkreten Anwendungsfall, dem SWR, geschaffen werden müssen.
Anschließend wird in einer experimentellen Phase im SWR-eigenen Netzwerklabor eine beispielhafte Architektur mit allen benötigten Komponenten zu Testzwecken konzeptioniert und implementiert. Dabei ist das Design und die Konfiguration möglichst nahe an der realen Umgebung des SWRs orientiert. Mit dem vorliegenden Testsetup werden dann bestimmte Funktionen und Anwendungsszenarien genauer aufgezeigt und die Relevanz für den Sicherheitsgewinn im SWR-LAN untersucht.
Darauf folgt eine abschließende Beurteilung des Sicherheits- und Effizienzgewinns durch die Einführung von SD-Access im SWR-Netzwerk.
This work addresses the conceptualization, design, and implementation of an Application Programming Interface (API) for the Common Security Advisory Framework (CSAF) 2.0, introducing another method for distributing CSAF documents in addition to two already existing methods. These don't allow the use of flexible queries as well as filtering, which makes it difficult for operators of software and hardware to use CSAF. An API is intended to simplify this process and thus advance the automation goal of CSAF.
First, it is evaluated whether the current standard allows the implementation of an API. Any conflicts are highlighted and suggestions for standard adaptations are made. Based on these results, the API is designed to meet the previously defined requirements. Subsequently, a proof of concept is successfully developed according to the design and extensively tested with specially prepared test data. Finally, the results and the necessary standard adjustments are summarized and justified.
The conceptual design and the implementation were successfully completed. However, during the implementation of the proof of concept, some routes could not be fully implemented.
Among the billions of smartphone users in the world, Android still holds more than 80% of the market share. The applications which the users install have a specific set of features that need access to some device functionalities and sensors that may hold sensitive information about the user. Therefore, Android releases have set permission standards to let the user know what information is being disclosed to the application. Along with other security and privacy improvements, significant changes to the permission scheme are introduced with the Android 6.0 version (API level 23). In this master thesis, the Android permission scheme is tested on two devices from different eras. The evolution of Android over the years is examined in terms of confidentiality. For each device, two applications are built; one focused on extracting every piece of information within the confidentiality scope with every permission declared and/or requested, and the other app focused on getting this type of information without user notification. The resulting analysis illustrates whether how and in what way the Android permission scheme declined or improved over time.
Threat Modeling is a vital approach to implementing ”Security by Design” because it enables the discovery of vulnerabilities and mitigation of threats during the early stage of the Software Development Life Cycle as opposed to later on when they will be more expensive to fix. This thesis makes a review of the current threat Modeling approaches, methods, and tools. It then creates a meta-model adaptation of a fictitious cloud-based shop application which is tested using STRIDE and PASTA to check for vulnerabilities, weaknesses, and impact risk. The Analysis is done using Microsoft Threat Modeling Tool and IriusRisk. Finally, an evaluation of the results is made to ascertain the effectiveness of the processes involved with highlights of the challenges in threat modeling and recommendations on how security developers can make improvements.
In the work at hand, we state that privacy and malleability of data are two aspects highly desired but not easy to associate. On the one hand, we are trying to shape data to make them usable and editable in an intelligible way, namely without losing their initial information. On the other hand, we are looking for effective privacy on data such that no external or non-authorized party could learn about their content. In such a way, we get overlapping requirements by pursuing different goals; it is trivial to be malleable without being secure, and vice versa. We propose four “real-world” use cases identified as scenarios where these two contradictory features are required and taking place in distinct environments. These considered backgrounds consist of firstly, cloud security auditing, then privacy of mobile network users and industry 4.0 and finally, privacy of COVID-19 tracing app users. After presenting useful background material, we propose to employ multiple approaches to design solutions to solve the use cases. We combine homomorphic encryption with searchable encryption and private information retrieval protocol to build an effective construction for the could auditing use case. As a second step, we develop an algorithm to generate the appropriate parameters to use the somewhat homomorphic encryption scheme by considering correctness, performance and security of the respective application. Finally, we propose an alternative use of Bloom filter data structure by adding an HMAC function to allow an outsourced third party to perform set relations in a private manner. By analyzing the overlapping bits occurring on Bloom filters while testing the inclusiveness or disjointness of the sets, we show how these functions maintain privacy and allow operations directly computed on the data structure. Then, we show how these constructions could be applied to the four selected use cases. Our obtained solutions have been implemented and we provide promising results that validate their efficiency and thus relevancy.
Das tiefe Lernen und die daraus entstehenden Technologien bieten eine neue Herausforderung für Unternehmen und privat Personen beiderlei. Deepfakes sind schon seit über vier Jahren im Internet verbreitet und in dieser Zeit wurden hauptsächlich politische Figuren Opfer der Technologie. Diese Arbeit nimmt sich das Ziel, die möglichen Attacken zu beschreiben und Gegenmaßnahmen dafür vorzustellen. Es wird zunächst Social Engineering erläutert und die technischen Grundlagen von Deepfakes gelegt. Daraufhin folgt ein Fallbeispiel, welches genauer aufzeigt, wie auch Unternehmen Opfer von Deepfake Attacken werden können. Diese Attacken fügen einen erheblichen finanziellen sowie Reputationsschaden an. Daher müssen verschiedene technische und organisatorische Maßnahmen gegenüber Deepfakes im Social Engineering Umfeld eingeführt werden. Durch die ständige Entwicklung der Technik werden diese Attacken in der Zukunft an Komplexität und Häufigkeit zunehmen. Unternehmen, Forscher und IT-Sicherheitsspezialisten müssen daher die Entwicklung dieser Attacken beobachten.
The identification of vulnerabilities is an important element of the software development process to ensure the security of software. Vulnerability identification based on the source code is a well studied field. To find vulnerabilities on the basis of a binary executable without the corresponding source code is more challenging. Recent research has shown how such detection can be performed statically and thus runtime efficiently by using deep learning methods for certain types of vulnerabilities.
This thesis aims to examine to what extent this identification can be applied sufficiently for a variety of vulnerabilities. Therefore, a supervised deep learning approach using recurrent neural networks for the application of vulnerability detection based on binary executables is used. For this purpose, a dataset with 50,651 samples of 23 different vulnerabilities in the form of a standardised LLVM Intermediate Representation was prepared. The vectorised features of a Word2Vec model were then used to train different variations of three basic architectures of recurrent neural networks (GRU, LSTM, SRNN). For this purpose, a binary classification was trained for the presence of an arbitrary vulnerability, and a multi-class model was trained for the identification of the exact vulnerability, which achieved an out-of-sample accuracy of 88% and 77%, respectively. Differences in the detection of different vulnerabilities were also observed, with non-vulnerable samples being detected with a particularly high precision of over 98%. Thus, the methodology presented allows an accurate detection of vulnerabilities, as well as a strong limitation of the analysis scope for further analysis steps.
With many advances in sensor technology and the Internet of Things, Vehicle Ad Hoc Net- work (VANET) is becoming a new generation. VANET’s current technical challenges are deploying decentralized architecture and protecting privacy. Because Blockchain features are decentralized, distributed, mass storage, and non-manipulation features, this paper designs a new decentralized architecture using Blockchain technology called Blockchain-based VANET. Blockchain-based VANET can effectively resolve centralized problems and mutual distrust between VANET units. To achieve this, it is needed to provide scalability on the blockchain to run for VANET. In this system, our focus is on the reliability of incoming messages on the network. Vehicles check the validity of the received messages using the proposed Bayesian formula for trust management system and some information saved in the Blockchain. Then, based on the validation result, the vehicle computes a rate for each message type and message source vehicle. Vehicles upload the computed rates to Roadside Units (RSUs) in order to calculate the net reliability value. Finally, RSUs using a sharding consensus mechanism generate blocks, including the net reliability value as a transaction. In this system, all RSUs collaboratively maintain the latest updated Blockchain. Our experimental results show that the proposed system is effective, scalable and dependable in data gathering, computing, organization, and retrieval of trust values in VANET.
In recent years, both the Internet of Things (IoT) and blockchain technologies have been highly influential and revolutionary. IoT enables companies to embrace Industry 4.0, the Fourth Industrial Revolution, which benefits from communication and connectivity to reduce cost and to increase productivity through sensor-based autonomy. These automated systems can be further refined with smart contracts that are executed within a blockchain, thereby increasing transparency through continuous and indisputable logging. Ideally, the level of security for these IoT devices shall be very high, as they are specifically designed for this autonomous and networked environment. This paper discusses a use case of a company with legacy devices that wants to benefit from the features and functionality of blockchain technology. In particular, the implications of retrofit solutions are analyzed. The use of the BISS:4.0 platform is proposed as the underlying infrastructure. BISS:4.0 is
intended to integrate the blockchain technologies into existing enterprise environments. Furthermore, a security analysis of IoT and blockchain present attacks and countermeasures are presented that are identified and applied to the mentioned use case.
In this work, an implementation of the somewhat homomorphic BV encryption scheme is presented. During the implementation, care was taken to ensure that the resulting program will be as efficient as possible i.e. fast and resource-saving. The basis for this is the work of Arndt Bieberstein, who implemented the BV scheme with respect to functionality. The presented implementation supports the basics of the BV scheme, namely (symmetric and asymmetric) encryption, decryption and evaluation of addition as well as multiplication. Additionally, it supports the encoding of positive and negative numbers, various gaussian sampling methods, basically infinitely large polynomial coefficients, the generation of suitable parameters for a use case, threading and relinearization to reduce the size of a ciphertext after multiplications. After presenting the techniques used in the implementation, it’s actual efficiency is determined by measuring the timings of the operations for various parameters.
In the field of network security, the detection of possible intrusions is an important task to prevent and analyse attacks. Machine learning has been adopted as a particular supporting technique over the last years. However, the majority of related published work uses post mortem log files and fails to address the required real-time capabilities of network data feature extraction and machine learning based analysis [1-5]. We introduce the network feature extractor library FEX, which is designed to allow real-time feature extraction of network data. This library incorporates 83 statistical features based on reassembled data flows. The introduced Cython implementation allows processing individual packets within 4.58 microseconds. Based on the features extracted by FEX, existing intrusion detection machine learning models were examined with respect to their real-time capabilities. An identified Decision-Tree Classifier model was thus further optimised by transpiling it into C Code. This reduced the prediction time of a single sample to 3.96 microseconds on average. Based on the feature extractor and the improved machine learning model an IDS system was implemented which supports a data throughput between 63.7 Mbit/s and 2.5 Gbit/s making it a suitable candidate for a real-time, machine-learning based IDS.
Die Nutzung Mobiler Endgeräte nimmt zu, Ortsgebundenheit immer mehr ab. Dieser Wandel bringt Veränderungen mit sich, so auch im Bereich Softwareinstallation. Das Installieren von Software über die Cloud ist im Konsumerbereich bereits stark vertreten. Nicht nur am Desktop, sondern insbesondere bei mobilen Endgeräte, wie Smartphones und Tablets wird Software, bzw. werden Apps fast ausschließlich nur noch über das Internet installiert.
Das ist ein aktuelles Thema für den Sensorhersteller SICK.
SICK entwickelt Sensoren, die programmierbar sind, sogenannte AppSpace-Sensoren, was über das Aufspielen von sogenannten SensorApps bereitgestellt wird.
Aufgrund dieser Entwicklung, dem Trend hin zu Cloudlösungen und Ortsungebundenheit, sieht das Unternehmen SICK die Notwendigkeit den Installationsprozess mithilfe einer mobilen App über das Internet anbieten zu können. Deshalb wurde im Rahmen der Arbeit eine hybride mobile App entwickelt, mit der es möglich ist, SensorApps über die Cloud auf ein SICK AppSpace-Gerät aufzuspielen. Dafür wurde auf Basis einer Nutzerrecherche ein Konzept entwickelt, welches im Anschluss mithilfe von gängigen Webtechnologien als hybride App umgesetzt wurde.
Möglichkeiten für die Verbesserung der Sicherheit von Endgeräten und Netzwerkstrukturen in der Cloud
(2021)
Meine Thesis soll sich mit dem Betrieb eines Unternehmens-Netzes mit Microsoft Tools mit und ohne Cloud beschäftigen. Dabei soll herausgearbeitet werden, in welchen Bereichen es Sinn ergibt, eine Cloudstrategie anzustreben und wo Unternehmen vielleicht empfohlen werden sollte, vorerst oder vielleicht auch langfristig bei der on-premise Variante zu bleiben.
Dafür werden Dinge wie rechtliche Aspekte, Datenschutz oder Kostenpunkte eher sekundär behandelt, da das Hauptaugenmerk auf der Sicherheit der Unternehmensdaten, Endgeräte und Server liegen soll.
Tatsächlich verfolgen bereits einige große Namen in Deutschland eine sogenannte "Cloud-first" Strategie, bei welcher versucht wird, alle möglichen Anwendungen in der Cloud unterzubringen. Auf Beweggründe und Motivation dieser Unternehmen wird kurz eingegangen.
Dann werden verschiedenste Bereiche behandelt, die im täglichen Unternehmensumfeld eine Rolle spielen, darunter fallen beispielsweise die Nutzung von Fileservern im Vergleich mit dem Cloud Ansatz OneDrive. Es soll die Frage beantwortet werden, ob eine vollständige Migration in die Cloud für verschiedene Bereiche Sinn ergibt, wie eben zum Beispiel den kompletten FileServer in die Cloud (im speziellen Fall OneDrive) umzuziehen.
Ein weiterer Unterpunkt ist die Nutzung eines Active Directorys bzw. von Domain Controllern in der Cloud über Azure AD. Hier wird insbesondere auch auf die Frage eingegangen, welche Unternehmen immer wieder beschäftigt, ob die, für die Konfiguration von Endgeräten sehr wichtigen, Group Policies in der Cloud erhalten bleiben bzw. ob und durch was sie ersetzt werden, oder ob und wie man sie migrieren kann.
Migration ist an dieser Stelle ein gutes Stichwort, denn ebenfalls soll aufgezeigt werden, wie herausfordernd die Migration verschiedenster on-premise Strukturen in die Cloud ist, und ob überhaupt bei allen Dingen eine Migration möglich ist. Interessant wird es hier möglicherweise bei legacy-Anwendungen.
Neben weiteren Themen wie dem sicherheitskritischen Allow-listing und verschiedenen Security-Ansätzen, denen mit einer Cloud-Strategie bessere Möglichkeiten gegeben werden kann, soll geklärt werden, welche Zusatzherausforderungen, aber auch Chancen die Cloud bieten kann und wird.
17
Zusammengefasst soll am Ende klarer sein, für was sich die Microsoft Cloud Anwendungen momentan schon gut eignen, wo man abwägen muss, und in welchen Aspekten die Cloud sich zunächst noch weiterentwickeln muss, bevor sie sinnvoll und verlässlich nutzbar ist. Außerdem soll aufgeführt werden, in welchen Bereichen man in den nächsten Jahren, und vielleicht auch schon jetzt, kaum noch an der Cloud vorbeikommt, wenn man eine bestimmte Funktionalität, sei es nun eine Anwendung, die die Sicherheit des Unternehmens erhöht, oder eine die die Produktivität steigern soll, nutzen möchte.
Die Thesis soll also eine Möglichkeit für Unternehmen bieten, sich unabhängig über Cloudangebote, hier meist am Beispiel der Microsoft Cloud Services, zu informierten. Und außerdem einzuschätzen, wie komplex und realisierbar verschiedene Dinge sind.
Die Fortschritte in der künstlichen Intelligenz ermöglichen es uns komplexe Aufgaben mit weniger Aufwand und in kürzerer Zeit zu lösen. Daraus ergibt sich auch die Möglichkeit komplexe Daten effizienter zu verarbeiten.
Smarte Technologien, insbesondere Smartphones und Wearables, sind in den letzten Jahren immer mehr zu alltäglichen Massenmedien geworden und sind aus dem Alltag vieler Menschen nicht mehr wegzudenken. Dadurch ergeben sich für Angreifer neue Angriffsvektoren, die die Privatsphäre und Sicherheit der Daten der Nutzer gefährden.
Zielaspekt dieser Arbeit ist es zu zeigen, dass eingebaute Sensoren in smarten Technologien ein Sicherheitsrisiko mit sich bringen, welches mithilfe von Deep-Learning ausgenutzt werden kann.
Hierfür wurden zwei LSTM-Netze implementiert, welche in der Lage sind Rückschlüsse auf eingetippte Ziffern und Buchstaben zu ziehen. Damit wird demonstriert, dass Machine-learning-basiertes Keylogging mittels externen Sensordaten möglich ist und der Angriff durch das Machine-learning vereinfacht wird.
Diese Arbeit beschäftigt sich mit der Entwicklung einer E-Learning Anwendung zum Peer-to-Peer Algorithmus Content Addressable Network (CAN). Ein CAN ist eine verteilte Hashtabelle zur dezentralen Verwaltung von Daten in Form von Schlüssel-Wert Paaren. Zweck der Anwendung ist eine didaktisch sinnvolle Aufbereitung und Darstellung der grundlegenden Abläufe im CAN. Die Anwendung simuliert ein CAN und bietet ein grafisches Interface zur Interaktion. Die Anwendung soll unterstützend im Rahmen des Moduls Advanced Networking im Informatik Master an der Hochschule Offenburg eingesetzt werden.
In dem letzten Jahr, welches von der Corona Pandemie geprägt war, wurde das Arbeiten von zu Hause aus durch die Unternehmen stark vorangetrieben. Auch wenn dies zuerst nur als Übergangslösung gesehen wurde, scheint dieser Trend sich auch nach der Pandemie fortzusetzen. In den meisten Heimnetzwerken ist die IT-Sicherheit allerdings deutlich schwächer ausgebaut, als dies in Unternehmen ist.
Da die Umstellung in das Home-Office sehr schnell durchgeführt wurde, sehen sich viele Arbeitnehmer und Arbeitgeber mit dieser Situation überfordert, auch im Heimnetzwerk des Anwenders für eine geeignete Sicherheit zu sorgen. In dieser Ausarbeitung werden daher einzelne Angriffe auf IT-Systeme im Heimnetzwerk aufgegriffen sowie Lösungsansätze, um eine sichere Heimnetzwerkumgebung zu schaffen.
Bereichsübergreifender Einsatz von JavaScript – Aktueller Stand und exemplarische Implementierung
(2021)
Nahezu alle Websites nutzen die Programmiersprache JavaScript zur Darstellung von interaktiven Inhalten und zur Bereitstellung von komplexen Funktionalitäten. Seit ihren Anfängen im Jahr 1995 hat sich die Sprache nicht nur zum Standard in der Webentwicklung etabliert, sondern auch zu einer leistungsfähigen Mehrzweckprogrammiersprache weiterentwickelt.
Diese Arbeit befasst sich mit einer ausführlichen Darstellung der aktuellen Möglichkeiten, welche Ansätze sich durch die Weiterentwicklung JavaScripts zu einer Mehrzweckprogrammiersprache ergeben und wie sich diese heute umsetzen lassen. Anhand des intelligenten Schlüsselkastens „Smart Vault“ wird verdeutlicht, wie dieses Vorgehen praktisch realisiert und die Vorteile einer einzigen Programmiersprache angewendet werden können.
Es hat sich herausgestellt, dass sich JavaScript für Anwendungen unterschiedlicher Bereiche einsetzen lässt und darüber hinaus ein hohes Potenzial für weitere Entwicklungen, Verbesserungen und zusätzliche Einsatzgebiete besitzt. Es lassen sich nicht nur Websites, Web Server und Desktop Apps, sondern auch Mikrocontroller im Internet of Things konfigurieren und miteinander nutzen, ohne eine weitere Programmiersprache zu benötigen. Zahlreiche Bibliotheken und Frameworks machen es möglich, dass die Sprache verschiedene Anwendungen über ihre Einsatzgebiete hinweg miteinander verbindet.
Duplikaterkennung, -suche und -konsolidierung für Kunden- und Geschäftspartnerdaten, sog. „Identity Resolution“, ist die Voraussetzung für erfolgreiches Customer Relationship Management und Customer Experience Management, aber auch für das Risikomanagement zur Minimierung von Betrugsrisiken und Einhaltung regulatorischer Vorschriften und viele weitere Anwendungsfälle. Diese Systeme sind jedoch hochkomplex und müssen individuell an die kundenspezifischen Anforderungen angepasst werden. Der Einsatz lernbasierter Verfahren bietet großes Potenzial zur automatisierten Anpassung. In diesem Beitrag präsentieren wir für ein KMU praxisfähige, lernbasierte Verfahren zur automatischen Konfiguration von Business-Regeln in Duplikaterkennungssystemen. Dabei wurden für Fachanwender Möglichkeiten entwickelt, um beispielgetrieben das Match-System an individuelle Business-Regeln (u.a. Umzugserkennung, Sperrlistenabgleich) anzupassen und zu konfigurieren. Die entwickelten Verfahren wurden evaluiert und in einer prototypischen Lösung integriert. Wir konnten zeigen, dass unser Machine-Learning-Verfahren, die von einem Domainexperten erstellten Business-Regeln für das Duplikaterkennungssystem „identity“ verbessern konnte. Zudem konnte der hierzu erforderliche Zeitaufwand verkürzt werden.
Due to the rapidly increasing storage consumption worldwide, as well as the expectation of continuous availability of information, the complexity of administration in today’s data centers is growing permanently. Integrated techniques for monitoring hard disks can increase the reliability of storage systems. However, these techniques often lack intelligent data analysis to perform predictive maintenance. To solve this problem, machine learning algorithms can be used to detect potential failures in advance and prevent them. In this paper, an unsupervised model for predicting hard disk failures based on Isolation Forest is proposed. Consequently, a method is presented that can deal with the highly imbalanced datasets, as the experiment on the Backblaze benchmark dataset demonstrates.
The recent successes and wide spread application of compute intensive machine learning and data analytics methods have been boosting the usage of the Python programming language on HPC systems. While Python provides many advantages for the users, it has not been designed with a focus on multiuser environments or parallel programming - making it quite challenging to maintain stable and secure Python workflows on a HPC system. In this paper, we analyze the key problems induced by the usage of Python on HPC clusters and sketch appropriate workarounds for efficiently maintaining multi-user Python software environments, securing and restricting resources of Python jobs and containing Python processes, while focusing on Deep Learning applications running on GPU clusters.
The interaction between agents in multiagent-based control systems requires peer to peer communication between agents avoiding central control. The sensor nodes represent agents and produce measurement data every time step. The nodes exchange time series data by using the peer to peer network in order to calculate an aggregation function for solving a problem cooperatively. We investigate the aggregation process of averaging data for time series data of nodes in a peer to peer network by using the grouping algorithm of Cichon et al. 2018. Nodes communicate whether data is new and map data values according to their sizes into a histogram. This map message consists of the subintervals and vectors for estimating the node joining and leaving the subinterval. At each time step, the nodes communicate with each other in synchronous rounds to exchange map messages until the network converges to a common map message. The node calculates the average value of time series data produced by all nodes in the network by using the histogram algorithm. The relative error for comparing the output of averaging time series data, and the ground truth of the average value in the network will decrease as the size of the network increases. We perform simulations which show that the approximate histograms method provides a reasonable approximation of time series data.
Seit 2009 nimmt das Team ”magmaOffenburg” an der 3D-Simulationsliga des RoboCups teil. Für das erfolgreiche Abschneiden in Turnieren ist die Qualität der erlernten Bewegungsabläufe ein zentraler Faktor. Bisher wurden genetische Algorithmen verwendet, um verschiedenste Aktionen zu erlernen sowie zu optimieren. In dieser Arbeit wird der Deep Reinforcement Learning Algorithmus Proximal Policy Optimization für das Erlernen bestimmter Bewegungen verwendet. Um ein Verständnis für dessen einflussreichen Parameter zu erhalten, werden Größen wie paralleles Lernen, Hyperparameter, Netzwerktopologie, Größe des Observationspace sowie asynchronem Lernen anhand dem Kicken aus dem Stand evaluiert. Durch die Ergebnisse der Evaluierung konnte der erlernte Kick signifikant verbessert werden und sein genetisch erlerntes Gegenstück im Spiel ablösen. Drüber hinaus wurden die Erkenntnisse anhand dem Laufen lernen evaluiert und Zusammenhänge bzw. Unterschiede der zwei Lernprobleme festgestellt.
The evolution of cellular networks from its first generation (1G) to its fourth generation (4G) was driven by the demand of user-centric downlink capacity also technically called Mobile Broad-Band (MBB). With its fifth generation (5G), Machine Type Communication (MTC) has been added into the target use cases and the upcoming generation of cellular networks is expected to support them. However, such support requires improvements in the existing technologies in terms of latency, reliability, energy efficiency, data rate, scalability, and capacity.
Originally, MTC was designed for low-bandwidth high-latency applications such as, environmental sensing, smart dustbin, etc. Nowadays there is an additional demand around applications with low-latency requirements. Among other well-known challenges for recent cellular networks such as data rate energy efficiency, reliability etc., latency is also not suitable for mission-critical applications such as real-time control of machines, autonomous driving, tactile Internet etc. Therefore, in the currently deployed cellular networks, there is a necessity to reduce the latency and increase the reliability offered by the networks to support use cases such as, cooperative autonomous driving or factory automation, that are grouped under the denomination Ultra-Reliable Low-Latency Communication (URLLC).
This thesis is primarily concerned with the latency into the Universal Terrestrial Radio Access Network (UTRAN) of cellular networks. The overall work is divided into five parts. The first part presents the state of the art for cellular networks. The second part contains a detailed overview of URLLC use cases and the requirements that must be fulfilled by the cellular networks to support them. The work in this thesis is done as part of a collaboration project between IRIMAS lab in Université de Haute-Alsace, France and Institute for Reliable Embedded Systems and Communication Electronics (ivESK) in Offenburg University of Applied Sciences, Germany. The selected use cases of URLLC are part of the research interests of both partner institutes. The third part presents a detailed study and evaluation of user- and control-plane latency mechanisms in current generation of cellular networks. The evaluation and analysis of these latencies, performed with the open-source ns-3 simulator, were conducted by exploring a broad range of parameters that include among others, traffic models, channel access parameters, realistic propagation models, and a broad set of cellular network protocol stack parameters. These simulations were performed with low-power, low-cost, and wide-range devices, commonly called IoT devices, and standardized for cellular networks. These devices use either LTE-M or Narrowband-IoT (NB-IoT) technologies that are designed for connected things. They differ mainly by the provided bandwidth and other additional characteristics such as coding scheme, device complexity, and so on.
The fourth part of this thesis shows a study, an implementation, and an evaluation of latency reduction techniques that target the different layers of the currently used Long Term Evolution (LTE) network protocol stack. These techniques based on Transmission Time Interval (TTI) reduction and Semi-Persistent Scheduling (SPS) methods are implemented into the ns-3 simulator and are evaluated through realistic simulations performed for a variety of low-latency use cases focused on industry automation and vehicular networking. For testing the proposed latency reduction techniques in cellular networks, since ns-3 does not support NB-IoT in its current release, an NB-IoT extension for LTE module was developed. This makes it possible to explore deployment limitations and issues.
In the last part of this thesis, a flexible deployment framework called Hybrid Scheduling and Flexible TTI for the proposed latency reduction techniques is presented, implemented and evaluated through realistic simulations. With help of the simulation evaluation, it is shown that the improved LTE network proposed and implemented in the simulator can support low-latency applications with low cost, higher range, and narrow bandwidth devices. The work in this thesis points out the potential improvement techniques, their deployment issues and paves the way towards the support for URLLC applications with upcoming cellular networks.
Bei der Produktion von Solarzellen aus multikristallinem Silizium haben Defekte aus der Kristallisationsphase starken Einfluss auf die Materialqualität der Wafer und damit auf den Wirkungsgrad der späteren Solarzelle. Ein Verständnis des Kornwachstums in multikristallinem Silizium während des Kristallisationsprozesses kann zur Optimierung desselben beitragen. In dieser Arbeit werden Methoden untersucht, optische Flüsse zwischen Korngrenzenbildern multikristalliner Si-Wafer mittels neuronaler Netze zu berechnen. Hierfür wird die Architektur eines ausgereiften faltungsbasierten neuronalen Netzes zur optischen Fluss-Berechnung genutzt und durch angepasstes Training auf Waferstrukturen zugeschnitten. Dies umfasst die Synthese eigener, auf Waferbilder basierender Trainingsdaten und das Training mit einer angepassten Fehlerfunktion zur Bewertung der Zuordnungsgenauigkeit von Körnern zwischen Wafern durch den optischen Fluss. Beide Maßnahmen zusammen führen zu einer Reduktion des Zuordnungsfehlers von Körnern zwischen Waferbildern um 45 % gegenüber einem hochoptimierten, auf allgemeine optische Flüsse trainierten Modell basierend auf demselben Netzwerk. Die geschätzte Zuordnungsgenauigkeit des besten Modells beträgt 92,4 % der Pixel der Korngrenzenbilder eines Wafers. Weiteres Verbesserungspotenzial ist vorhanden.
Annotated training data is essential for supervised learning methods. Human annotation is costly and laborsome especially if a dataset consists of hundreds of thousands of samples and annotators need to be hired. Crowdsourcing emerged as a solution that makes it easier to get access to large amounts of human annotators. Introducing paid external annotators however introduces malevolent annotations, both intentional and unintentional. Both forms of malevolent annotations have negative effects on further usage of the data and can be summarized as spam. This work explores different approaches to post-hoc detection of spamming users and which kinds of spam can be detected by them. A manual annotation checking process resulted in the creation of a small user spam dataset which is used in this thesis. Finally an outlook for future improvements of these approaches will be made.