With the rise of the artificial intelligence starting in the late 2010s there has been a great increase of demand for neural networks. It seems AI is omnipresent these days, whether it is OpenAI releasing the famous or infamous large language model chatGPT, Google following by adopting a similar approach with integrating Artificial Intelligence (AI) into it's search engine or help desks becoming entirely controlled by AI. It is clear, AI is here to stay. NVIDIA normally a company thought of selling Graphical Processing Units (GPUs) for gaming purposes has shifted its focus to making GPUs used for running neural networks, this shift has shown its returns as it is now the most valuable company in the world.

The increased interest in AI does leave a lot of academia wondering: What actually is artificial intelligence and especially what are neural networks. With the introduction of deep learning, a machine learning method that searches for features in a dataset, neural networks have become more and more abstract. Developing an appropriate solution to better interpret, visualize and adjust the inner workings of a neural network is not just needed to keep track of the underlying dynamics of a neural network, but also to allow the user to go outside of the standardized framework to fine tune a network.

This thesis explores the functionality of neural networks, more specifically spiking neural networks, and describes the workings and functions of the created simulator. Multiple helper functions, that allow the user to both visualize and adjust the network outside of the standardized framework, are introduced that distinguishes the created simulator from the already present SNN simulators. Afterwards, five small simulations are done to validate the basic functionality of the simulator. Following these simulations the simulator is validated using two large datasets and comparing the simulator with the results from a well known spiking neural network library SLAYER. The first validation is done with a smaller data set containing hand written images of the numbers zero to nine. The second validation is done with a larger data set containing multiple different kind of performed hand gestures. Both validations show that the simulator has a high accuracy with the maximum mean square error between the SLAYER output and the simulators output being less than 2%.

This thesis researches the security of firmware images in the Internet of Things (IoT) and embedded devices. We present an open-source tool, Embedded Binary Analysis Tool (EBAT), designed to analyze cross-architectural firmware image security context. EBAT consists of various modules capable of discovering outdated software for various libraries, particularly on cryptographic libraries, and detecting Common Vulnerabilities and Exposures (CVEs), focusing on firmware's cryptographic libraries. It also detects exploit mitigation techniques on firmware's image binaries and discovers credentials and passwords with a focus on private keys embedded in the firmware image. Additionally, EBAT identifies Application Programming Interfaces (APIs) cryptographic misuses through static taint analysis (backward tracking) on cross-architectural binaries. We presented a total of 18 well-defined cryptographic rules and a list of 733 function calls with more than 1,600 function arguments, applicable in static taint analysis to check the possibility of cryptographic misuses based on 10 well-used open-source cryptographic libraries APIs. EBAT's static taint analysis provides a powerful framework for detecting the possibility of cryptographic misuses in cross-architectural binaries, making it a valuable tool for identifying and addressing vulnerabilities in cryptographic implementation in firmware images.

Using EBAT, we conducted a large-scale analysis of over 36,000 firmware images publicly crawled from the Internet and successfully unpacked over 60% of them. The created dataset of firmware images includes more than 5,000 different products across 33 vendors, spanning more than 20 years and a plethora of various device types. Our findings show that ARM and MIPS are the most prevailed CPU architecture in the IoT/embedded industry. We compared identical binaries across all vendors, revealing a significant percentage of similar binaries used across different vendors' firmware images. Our analysis of firmware binaries reveals a notable absence of exploit mitigation techniques in IoT/embedded firmware images, and we present many firmware images containing private keys, posing potential security threats. Additionally, versions of open-source cryptographic libraries used in firmware images are identified, and the CVEs of the cryptographic libraries are evaluated. Two real-world case studies on hard-coded credentials demonstrate the significance of the large-scale attack presented in this thesis. Hashed passwords, predominantly using outdated algorithms, have also been discovered, and several have been cracked.

The main goal of EBAT is to identify cryptographic misuses in cross-architectural binaries. By applying static taint analysis (backward tracking) to well-defined APIs on specific functions and arguments for 10 open-source cryptographic libraries, we can identify potential violations of cryptographic rules. This analysis was executed on over 1.4 million binaries, revealing that approximately 50% of examined firmware images violated at least one cryptographic rule. Various case studies on real-world vulnerabilities in firmware images are presented, including recent CVEs that are found in various vendors' products. Executing EBAT on those vulnerable firmware images, we tested the effectiveness of our tool to evaluate the automatic capturing of these known vulnerabilities. In addition, performing large-scale analysis on an extensive corpus of firmware images allows us to discover that other firmware images are affected by these known vulnerabilities, in some cases also across various product lines not covered on the public CVEs reports.

In conclusion, EBAT is a valuable resource for researchers working on firmware security. Its automated analysis process, comprehensive modules, and ability to discover possible vulnerabilities, cryptographic misuses at a binary level, and other security weaknesses make it a powerful tool for identifying and mitigating security risks in IoT/embedded devices.

For my master’s thesis, I developed a novel indoor localization approach that can achieve accurate localization results. Indoor localization is a well-known topic of research, and many attempts have been made to find the so-called holy grail. For a robotic car swarm to be able to execute a specific task in a specific room, each robot in the swarm needs to know its own location with respect to a map of the building. Allowing the robot to deduce this location himself allows for total automation of the swarm and helps mitigate errors along the way due to faulty sensor readings. Numerous works have tried to solve this problem by using beacons or heavily trained machine learning networks. These methods, however, prevent the robot from working everywhere, as either the location needs to be adapted or the robot needs to be trained for the location. To mitigate these issues, the work of this thesis is focused on achieving indoor localization that works everywhere by researching if a combination of a particle filter and sound source localization techniques can achieve high-accuracy indoor localization.

To achieve this, particle filter and source localization techniques are combined based on probability theory. Here, so-called localization tables are used to estimate the robot’s position based on the geo- metrical properties of the map, the DOA, and the distance from a received message. This table is then used to update the position and weight of the particles, which can then be used to make an educated guess of the robot’s position. This approach proved to be effective, as it was able to achieve average RMSE results as low as 3.83 cm, where the robot only needed to drive 363 cm and localization was achieved within 9.52 seconds. The research showed that deploying more cars in the swarm leads to better results, as fewer transmissions are needed, and less distance needs to be traveled. In addition, the research also showed that localization can be achieved while driving even less distance by sharing the localization tables between the robots. This, however, does come at the cost of localization accu- racy, resulting in an RMSE of 19.89 cm while driving only 158 cm. To conclude this research, based on the results of the novel fusion of a particle filter and sound source localization techniques, it can be concluded that the work in this thesis offers a great contribution to the field of indoor localization using audio-based signals.

In sequential decision-making, Multi-armed Bandit (MAB) models the dilemma of exploration versus exploitation. The problem is commonly situated in an unknown environment where a player iteratively selects one action from a set of predetermined choices. The player's choices can be evaluated by comparing observed rewards after each decision to the highest one obtainable from the set of possible actions. The goal is to minimize the measure of regret of not choosing the optimal action every time. Intuitively, one might think of exploring all possible options first and then selecting the one for which the highest rewards were observed. However, it is difficult to model this behaviour when the specifications of the environment setting are not known or when it changes in time. Some classifications of environments exist and can be used to decide which approach can be used to model the MAB problem. The literature provides various algorithms used to model different MAB problems but lacks comparisons between proposed algorithms across the domains they aim to solve. To fill this gap, we want to focus on several Multi-armed Bandit algorithms (policies) and compare their effectiveness and optimality in different environments. This research focuses on a class of environments with a sparse reward function, where the reward depends on the action-specific contextual information and some sparse vector that dictates which features are considered. Four MAB algorithms were selected, each developed to solve the problem in a different environment. The comparison was made in a series of experiments to explore how sparsity affects their performance. The experiments conclude that the advantage of using sparsity-adopted algorithms depends on both sparsity and the environment setting, but increasing the sparsity helps achieve better results than traditional methods.

This thesis investigates the performance of various bandit algorithms in non-stationary contextual environments, where reward functions change unpredictably over time. Traditional bandit algorithms, designed for stationary settings, often fail in dynamic real-world scenarios. This research evaluated the adaptability and computational performance of popular algorithms such as UCB, LinUCB, and LinEXP3 using a self-implemented bandit framework. Empirical results reveal significant insights into the trade-offs and optimal strategies for applying these algorithms in non-stationary conditions. Notably, LinEXP3 demonstrated superior performance in complex environments due to its ability to incorporate Bayesian posteriors, despite its higher computational cost. The key contributions of this paper include the empirical evaluation of these algorithms and their implementations, with tailored environment settings. The results suggest promising directions for further research, including the incorporation of broader algorithmic ranges like Contextual Thompson Sampling and other reinforcement learning algorithms adapted for linear contextual settings. Additionally, future work should focus on using real-world datasets to validate these algorithms and introducing covariance matrices for context vectors to simulate more realistic learning processes. These findings could influence the design and implementation of bandit algorithms in practical applications such as recommendation systems and financial portfolio management.

The aim of this paper is to show experimental data on the regret-based performance of various solver algorithms within a class of decision problems called Multi-Armed Bandits. This can help to more efficiently choose the algorithm most suited for an application and to reduce the amount of modification work needed to adapt the algorithm to fulfill its purpose. This is done because current research is largely based on theoretical analysis, strictly within a specific environment, using random data. These studies can be difficult to use to predict the performance of a practical implementation in a less strictly defined environment, but the amount of data on more directly comparable practical implementations is limited. The experiments involve several of the environments which some of the algorithms have been designed for, including a basic stochastic environment, a static linear contextual environment and two different changing contextual environments. Based on the experiments ran in these environments, we conclude that algorithms designed for stochastic environments tend to perform a lot worse in contextual environments than algorithms designed for linear contextual environments, and that all of the algorithms we tested that were designed for contextual environments tend to perform similarly in most cases. We also find that in most cases the contextual algorithms’ performance relative to each other in changing environments is generally similar to their relative performance in static ones, even though the absolute regret behaviour tend to be different in the different environments. Our findings also include various more specific strengths and weaknesses for the individual algorithms.

Delay is a frequently encountered phenomenon in Multi-armed bandit problems that affects the accuracy of choosing the optimal arm. One example of this phenomenon is online shopping, where there is a delay between a user being recommended a product and placing the order. This study investigates the influence of delay in contextual multi-armed bandit settings, focusing on the performance of the OTFLinUCB algorithm, an adaptation of the Lin- UCB algorithm designed to handle delayed feed- back. The cumulative regret associated with various bandit algorithms, including UCB, Exp3, Lin- UCB, and OTFLinUCB, is examined under different delayed environments. Experiments con- ducted using artificial data generated to simulate real-world scenarios reveal that while delay cannot be entirely mitigated, its impact can be minimized through the strategic selection of hyperparameters, particularly the time window size for OTFLinUCB. The findings indicate that OTFLinUCB’s performance is highly dependent on the chosen window size, which balances the trade-off between estimator accuracy and memory consumption. A method for determining the optimal window size is proposed by introducing an intermediary distribution- agnostic term, conversion rate. Empirical evidence suggests that maintaining a conversion rate of around 70% achieves a balance between minimizing cumulative regret and optimizing memory usage. As a result, this research contributes to the understanding of the delays’ effect on bandit algorithms in contextual environments and offers practical guidelines for tuning OTFLinUCB’s time window.

The aim of this paper is to challenge and compare several Multi-Armed Bandit algorithms in an en- vironment with fixed kernelized reward and noisy observations. Bandit algorithms are a class of decision-making problems with the goal of opti- mizing the trade-off between exploration and ex- ploitation of all choices. Each decision yields some reward, and the goal is to minimize the regret that follows from a combination of decisions, that is to say to minimize the difference between the set of decisions made, and the set of optimal decisions. In particular, these algorithms deal with the trade- off between choosing the best-known option and exploring new, possibly better options. These al- gorithms are widely used in reinforcement learn- ing, optimization and economics, where decisions need to be made without all the information and with some uncertainty. Each environment is dif- ferent however, and some algorithms are better in some environments than others.

Saxophone is a musical instrument that tends to be out of tune. I have invented an automatic tuning saxophone mouthpiece that can help saxophonists play in tune by correcting their notes' frequencies in real time during performances using a stepper motor.

This thesis addresses the challenge of refactoring untyped actor systems into typed ones, particularly within the Scala ecosystem using Akka framework \cite{akkaTypedDocs/Online}. The actor model, with its message-passing architecture, offers a solution to concurrency and scalability issues in distributed systems, but transitioning from untyped Akka Classic to typed Akka actors is a complex and error-prone task. This work proposes a solution through the development of a communication flow graph that captures actor interactions and message exchanges, which is then used to automate the refactoring process while preserving system behavior and communication integrity.

Key contributions include the implementation of the communication flow graph and an automated refactoring tool that translates untyped actor systems to typed ones. The evaluation demonstrates that while the refactored systems maintain communication patterns and functionality, they may introduce a slight reduction in maintainability. Additionally, the resulting effectiveness of the refactored benchmarks varies depending on the complexity of the underlying system.

Lastly, this research sets the foundation for improving refactoring tooling aimed at actor systems, extending the usability of the communication flow graph for the automated documentation of actor systems, and highlights the need to revisit traditional software quality metrics to suit actor-based systems.

Internet of Things (IoT) devices regularly process sensitive data, including personal information. Therefore, ensuring their security is crucial to avoid damage and prevent data breaches. The Advanced Encryption Standard (AES) is generally regarded as one of the most popular cryptographic algorithms for ensuring data security. Typical lightweight implementations of the algorithm published in the literature focus on area and power optimization, while neglecting the performance. This paper presents a novel lightweight approach for the AES algorithm and considers both encryption and decryption. In terms of performance per unit area and performance per unit power, our 32-bit design outperforms the state-of-the-art by 1.69x and 1.27x, respectively. These improvements become even larger when implementing higher data-path designs, such as 64-bit or 128-bit designs. Our non-DOM AES design is secure against Correlation Power Analysis (CPA) but vulnerable to Template Based Attack (TBA) when more than 1500 traces are considered. To enhance its resilience against side channel attacks (SCAs), we modified our design by adopting and further improving on the most recent countermeasure, i.e., Domain-Oriented Masking (DOM). The results demonstrate that incorporating DOM into our design enables it to withstand against both CPA and TBA. Besides, our simplified eight-stage and five-stage 1st-order DOM SBOX designs achieve a reduction in area of 9.9% and 6.9% compared to the original proposed designs, respectively.

Rowhammer is a security exploit used to cause bit errors DRAM chips. Newer DRAM technologies are becoming more vulnerable to rowhammer attacks, and existing protec- tion methods are starting to reach their limits. This thesis provides methods for DRAM characterization by means of reverse engineering, an in-depth analysis of vulnerable cells during rowhammer and a novel protection method. First the DRAM module is characterized by performing a retention test and a one- sided rowhammer attack throughout the memory. The resilience to rowhammer of vul- nerable victim cells has been analysed by using various data patterns in the victim cell itself in combination with its direct and diagonal neighbours. The impact of the pro- posed protection method is measured by flipping the data in the attacker row during a simulated rowhammer attack. Various rowhammer sequences are investigated. The results following the DRAM characterization show that only a one-sided rowham- mer attack is possible. Rows that impact one another during rowhammer come in pairs of two, one attacker row impacts only one neighbouring row. The results of the in-depth analysis of cells during rowhammer shows evidence of negative horizontal impact com- ing from uncharged neighbouring cells in the same row and negative diagonal impact coming from charged cells neighbouring the attacker cell in the attacker row. The hori- zontal impact is mirrored in symmetrical cells whilst the diagonal impact is not. Results following the proposed protection method experiments show that using the protection method improves the overall resilience to rowhammer ranging from 50%, 65% to 100% depending on the hammer sequence.

This paper describes the feasibility of digit detection using three photodiodes and an Arduino Nano 33 BLE. This is done using a controlled lighting condition, using a bright lamp. It dives into the process of data collection, preprocessing and model selection for a recurrent neural network to do the classification of gestures into digits. Using a ”ConvLSTM double conv. layer model” with 128 units we were able to achieve an average accuracy of 0.500 ± 0.091 on a 5-Fold cross-validation procedure based on data that was collected in a controlled lighting environment. While this provides a foundation for digit detection using time-series data in a controlled light environment, further suggestions are made for future improvements and expansions in this area.

This paper presents a study focused on developing an efficient signal processing pipeline and identifying suitable machine learning models for real-time gesture recognition using a testbed consisting of an Arduino Nano 33 BLE and three OPT101 photodiodes. Our research aims to address the challenges of limited computational power whilst maintaining a high inference accuracy.

Experiments were conducted to optimise the signal processing and explore various machine learning model architectures, specifically revolving around convolutional neural networks. The data used for these experiments was gathered by creating a dataset of gestures from left- and right-handed participants. We took ethical considerations regarding participant recruitment and data security into account and we made sure to balance the dataset with both left- and right-handed participants as much as possible.

We obtained accurate gesture recognition results, surpassing the goal of a 75% success rate. Our machine learning models, trained on pre-processed 2D data, achieved near real-time inference times while running on the resource-constrained Arduino Nano 33 BLE.

The findings of this study contribute to the field of gesture recognition by providing insights into efficient signal processing techniques and identifying suitable machine learning models for resource-constrained devices. The developed system can be applied in various applications, ranging from games to healthcare. Furthermore, a dataset is contributed which can be used for further research.

Visible light sensing is a field of research that creates new possibilities for human-computer interaction. This research shows the viability of designing a system for detecting hand gestures using a cost-effective detection circuit employing 3 light-sensitive photodiodes. The way this research shows viability is by developing a machine-learning model that works on 3D-structured sensor data that is able to distinguish 10 different gestures and deploying the model on a standalone Arduino Nano 33 BLE microcontroller controlling the system. Using a combination of Convolutional Neural Networks and Recurrent Neural Networks it is possible to deploy a model called ConvLSTM-128 that achieves an accuracy of 70% on a dataset of limited size. This research acknowledges that the achieved accuracy is not suitable for real-world use, but concludes by outlining steps that could help future research in increasing the accuracy. Furthermore, an analysis of the 10 gestures shows that in order to improve accuracy, the way some gestures are performed might need alteration. Finally, a model size of around 140Kb and an inference time of 660ms show that this model is compact and fast enough to be deployed in real-world applications.

Touchless interaction with computers has become more important in recent years, especially in the context of the COVID-19 pandemic.
Applications include situations where touch input is not possible or not desirable, e.g. for hygienic purposes in a public setting or a medical setting.
Practical examples for touchless interaction include elevators, vending machines, and other public devices that are used by many people.
However, most touchless interaction systems are expensive and require significant computational power.
This paper proposes a bare-bones low-power and low-cost system for recognizing air-written digits using a microcontroller and light sensitivity sensors.
A proof of concept has been created and tested in a fixed lighting scenario with a fixed set of gestures for the digits 0 to 9 to show the feasibility of such a system.
The system uses a convolutional neural network to recognize digits and achieves an average accuracy of 58,8% on a validation set of unseen participants.
It performs significantly better on new samples from users already seen during training, achieving an average accuracy of 93,5%.

This thesis has studied an oversight in the scientific literature on Wi-Fi 6 modelling and attempts to quantify the prevalence and effect of this omission.
This effect was dubbed the ”backoff reset” and roughly means that a device connected to a Wi-Fi 6 network that recently received the green-light to send from the access point and, as such, has successfully sent a message, should reset the remaining time it will wait to send its next message.
The problem statement was then based on this effect and read as follows:
”What is the effect on performance metrics of resetting a station’s backoff stage and counter after an UL-OFDMA transmission?”
To allow readers new to the Wi-Fi space to be able to follow along, a brief introduction was given to some of the core concepts of Wi-Fi 6 and a preliminary literature survey was performed to find out how often the backoff reset is omitted.
From this census, it was found that only five out of fifteen papers likely included the backoff reset in their results.
To analyze the effects of the backoff reset on network performance, a series of experiments were performed in a simulator.
In these experiments, the system with a backoff reset, named the ”reset” system,
was pitched against a system without the reset, called the ”continue” system, while varying various network parameters.
From these results, the following conclusion is drawn: ”The ”reset” system outperforms the ”continue” system in certain situations on (a combination of) latency and collision probability, but is generally strictly worse on all five key metrics in a typical, flexible Wi-Fi deployment, where data rates and the amount of
connected devices can change rapidly. We thus accept the research hypothesis.” Finally, an overview of the situations in which the ”reset” system can outperform the ”continue” system was given and the limitations of this study were discussed, after which future work was proposed.

The current state of teleoperation in Tactile Internet faces the problem of limited operating distance. To circumvent this limitation, a setup has been proposed which provides Ultra-Low Latency feedback to the operator from a local simulation of the remote environment. To be able to accurately simulate the remote environment and the objects within it, an initial estimation of the mass and Center of Mass of these objects is required. We explore three approaches to the estimation of these properties, with the use of Axis-Aligned Bounding Bounding Boxes (AABB), Oriented Bounding Boxes (OBB), and convex hulls. We also explore the impact of sensor resolution and object occlusion. The performance of these methods varies, and highly depends on the type of object. Sensor resolution has a high impact on the estimated mass, but not on Center of Mass. Object occlusion has a clear impact on both, suggesting the use of multiple view points being will improve accuracy.