Refine
Document Type
- Master's Thesis (34) (remove)
Has Fulltext
- yes (34)
Is part of the Bibliography
- no (34)
Keywords
- IT-Sicherheit (4)
- Deep learning (3)
- Energiemanagement (2)
- Landwirtschaft (2)
- Maschinelles Lernen (2)
- 5G (1)
- Agrophotovoltaics (1)
- Algorithmus (1)
- Android (1)
- Attention-basiert (1)
Institute
- Fakultät Elektrotechnik, Medizintechnik und Informatik (EMI) (ab 04/2019) (34) (remove)
Open Access
- Closed Access (17)
- Closed (13)
- Open Access (4)
- Diamond (3)
This thesis focuses on the development and implementation of a Datagram Transport Layer Security (DTLS) communication framework within the ns-3 network simulator, specifically targeting the LoRaWAN model network. The primary aim is to analyse the behaviour and performance of DTLS protocols across different network conditions within a LoRaWAN context. The key aspects of this work include the following.
Utilization of ns-3: This thesis leverages ns-3’s capabilities as a powerful discrete event network simulator. This platform enables the emulation of diverse network environments, characterized by varying levels of latency, packet loss, and bandwidth constraints.
Emulation of Network Challenges: The framework specifically addresses unique challenges posed by certain network configurations, such as duty cycle limitations. These constraints, which limit the time allocated for data transmission by each device, are crucial in understanding the real-world performance of DTLS protocols.
Testing in Multi-client-server Scenarios: A significant feature of this framework is its ability to test DTLS performance in complex scenarios involving multiple clients and servers. This is vital for assessing the behaviour of a protocol under realistic network conditions.
Realistic Environment Simulation: By simulating challenging network conditions, such as congestion, limited bandwidth, and resource constraints, the framework provides a realistic environment for thorough evaluation. This allows for a comprehensive analysis of DTLS in terms of security, performance, and scalability.
Overall, this thesis contributes to a deeper understanding of DTLS protocols by providing a robust tool for their evaluation under various and challenging network conditions.
AI-based Ground Penetrating Radar Signal Processing for Thickness Estimation of Subsurface Layers
(2023)
This thesis focuses on the estimation of subsurface layer thickness using Ground Penetrating Radar (GPR) A-scan and B-scan data through the application of neural networks. The objective is to develop accurate models capable of estimating the thickness of up to two subsurface layers.
Two different approaches are explored for processing the A-scan data. In the first approach, A-scans are compressed using Principal Component Analysis (PCA), and a regression feedforward neural network is employed to estimate the layers’ thicknesses. The second approach utilizes a regression one-dimensional Convolutional Neural Network (1-D CNN) for the same purpose. Comparative analysis reveals that the second approach yields superior results in terms of accuracy.
Subsequently, the proposed 1-D CNN architecture is adapted and evaluated for Step Frequency Continuous Wave (SFCW) radar, expanding its applicability to this type of radar system. The effectiveness of the proposed network in estimating subsurface layer thickness for SFCW radar is demonstrated.
Furthermore, the thesis investigates the utilization of GPR B-scan images as input data for subsurface layer thickness estimation. A regression CNN is employed for this purpose, although the results achieved are not as promising as those obtained with the 1-D CNN using A-scan data. This disparity is attributed to the limited availability of B-scan data, as B-scan generation is a resource-intensive process.
Go ist eine 2009 veröffentlichte Programmiersprache mit einem statischen Typsystem. Seit Version 1.18 sind auch Generics ein Teil der Sprache. Deren Übersetzung wurde im de facto Standard-Compiler mittels Monomorphisierung umgesetzt. Diese bringt neben einigen Vorteilen auch Nachteile mit sich. Aus diesem Grund beschäftigt sich diese Arbeit mit einer alternativen Übersetzungsstrategie für Generics in Go und implementiert diese in einem neuen Compiler für Featherweight Generic Go, einem Subset von Go. Zum Schluss steht damit ein nahezu funktionierender Compiler, welcher schließlich Racket-Code ausgibt. Eine Evaluierung der Performanz der Übersetzungsstrategie ist allerdings noch ausstehend.
Hohe Kosten bei der Annotation von Daten führen dazu, dass datensparsamere Wege zum Erstellen von Modellen gesucht werden. In dieser Arbeit wird ein Lösungsansatz untersucht, der ausgehend von fokussierten Repräsentationen, datensparsame Lösungen für verschiedene Aufgaben finden soll. Durch einen Multi-Task-Learning-Ansatz trägt das Finden einer Repräsentation gleichzeitig zum Lösen einer Aufgabe bei. Durch Ersetzung einer der Aufgaben werden Wissentransfers datensparsam auf die neue Aufgabe durchgeführt. In der erarbeiteten und evaluierten Lösung können Hyperparameter automatisch gefunden werden. Bei Vergleichen von verschiedenen Ansätzen und Datenmengen ist über die Leistung der Netzwerke zu erkennen, dass der Ansatz insbesondere mit weniger Daten bessere Ergebnisse erzielt. Die Ergebnisse dieser Arbeit lassen eine Bereitstellung als Module zu. Die Module werden im Rahmen dieser Arbeit beschrieben. Abgeschlossen wird die Arbeit mit einem Ausblick auf Verbesserungen und Potenziale der Ansätze.
In the past ten years, applications of artificial neural networks have changed dramatically. outperforming earlier predictions in domains like robotics, computer vision, natural language processing, healthcare, and finance. Future research and advancements in CNN architectures, Algorithms and applications are expected to revolutionize various industries and daily life further. Our task is to find current products that resemble the given product image and description. Deep learning-based automatic product identification is a multi-step process that starts with data collection and continues with model training, deployment, and continuous improvement. The caliber and variety of the dataset, the design selected, and ongoing testing and improvement all affect the model's effectiveness. We achieved 81.47% training accuracy and 72.43% validation accuracy for our combined text and image classification model. Additionally, we have discussed the outcomes from the other dataset and numerous methods for creating an appropriate model.
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.
Server Side Rendering (SSR), Single Page Application (SPA), and Static Site Generation (SSG) are the three most popular ways of making modern Web applications today. If we go deep into these processes, this can be helpful for the developers and clients. Developers benefit since they do not need to learn other programming languages and can instead utilize their own experience to build different kinds of Web applications; for example, a developer can use only JavaScript in the three approaches. On the other hand, clients can give their users a better experience.
This Master Thesis’s purpose was to compare these processes with a demo application for each and give users a solid understanding of which process they should follow. We discussed the step-by-step process of making three applications in the above mentioned categories. Then we compared those based on criteria such as performance, security, Search Engine Optimization, developer preference, learning curve, content and purpose of the Web, user interface, and user experience. It also talked about the technologies such as JavaScript, React, Node.js, and Next.js, and why and where to use them. The goals we specified before the program creation were fulfilled and can be validated by comparing the solutions we gave for user problems, which was the application’s primary purpose.
The goal of this thesis is to thoroughly investigate the concepts of stand-alone and decarbonization of optical fiber networks. Because of their dependability, fast speed, and capacity, optical fiber networks are vital inmodern telecommunications. Their considerable energy consumption and carbon emissions, on the other hand, constitute a danger to global sustainability objectives and must be addressed.
The first section of the thesis presents a summary of the current state of optical fiber networks, their
components, and the energy consumption connected with them. This part also goes over the difficulties of lowering energy usage and carbon emissions while preserving network performance and dependability.
The second section of the thesis focuses on the stand-alone idea, which entails powering the optical fiber network with renewable energy sources and energy-efficient technology. This section investigates and explores the possibilities of renewable energy sources like solar and wind power to power the network. It also investigates energy-efficient technologies like virtualization and cloud computing, as well as their potential to minimize network energy usage.
The third section of the thesis focuses on the notion of decarbonization, which entails lowering carbon emissions linked with the optical fiber network. This section looks at various carbon-reduction measures, such as employing low-carbon energy sources and improving energy efficiency. It also covers the relevance of carbon offsets and the difficulties associated with adopting decarbonization measures in the context of optical fiber networks.
The fourth section of the thesis compares the ideas of stand-alone and decarbonization. It investigates the advantages and disadvantages of each strategy, as well as their potential to minimize energy consumption and carbon emissions in optical fiber networks. It also explores the difficulties in applying these notions as well as potential hurdles to their wider adoption.
Finally, the need of addressing the energy consumption and carbon emissions connected with optical fiber networks is emphasized in this thesis.
It outlines important obstacles and potential impediments to adopting these initiatives and gives insights into potential ways for decreasing them.
It also makes suggestions for further study in this area.
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.
Decarbonisation Strategies in Energy Systems Modelling: APV and e-tractors as Flexibility Assets
(2023)
This work presents an analysis of the impact of introducing Agrophotovoltaic technologies and electric tractors into Germany’s energy system. Agrophotovoltaics involves installing photovoltaic systems in agricultural areas, allowing for dual usage of the land for both energy generation and food production. Electric tractors, which are agricultural machinery powered by electric motors, can also function as energy storage units, providing flexibility to the grid. The analysis includes a sensitivity study to understand how the availability of agricultural land influences Agrophotovoltaic investments, followed by the examination of various scenarios that involve converting diesel tractors to electric tractors. These scenarios are based on the current CO2 emission reduction targets set by the German Government, aiming for a 65% reduction below 1990 levels by 2030 and achieving zero emissions by 2045. The results indicate that approximately 3% of available agricultural land is necessary to establish a viable energy mix in Germany. Furthermore, the expansion of electric tractors tends to reduce the overall system costs and enhances the energy-cost-efficiency of Agrophotovoltaic investments.