Electrical and Electronic Engineering - Masters by Research Theses
Permanent URI for this collection
Browse
Recent Submissions
Item Cryo-CMOS modelling and calibration at 4 Kelvin: tools, methods, and applications(University College Cork, 2025) Montanares Sepúlveda, Mauricio; McCarthy, Kevin G.; Salgado, GerardoQuantum computing has become increasingly significant due to its potential to efficiently address complex problems that classical computing struggles to solve. A primary obstacle to practical quantum computing implementations is the scalable integration and precise control of numerous quantum bits (qubits). Solid-state qubits, such as spin-based and superconducting types, require milli-Kelvin temperatures to maintain quantum coherence and proper operation. Current qubit-control methodologies employing room-temperature electronics introduce considerable thermal loads and wiring complexity, limiting scalability. Cryogenic CMOS (Cryo-CMOS), which utilises standard CMOS technology operating at 4 K—near qubit temperatures—represents a promising solution to these challenges. Nonetheless, standard Process Design Kits (PDKs), calibrated for conventional temperature ranges (–55 °C to 125 °C), fail to accurately capture significant changes in semiconductor behaviour under cryogenic conditions, including increased threshold voltages, enhanced carrier mobility, and altered subthreshold characteristics. The successful implementation of Cryo-CMOS controllers thus relies on precise cryogenic device modelling and calibration techniques. Although commercial PDK-calibration tools exist, they are typically expensive, with essential modules for advanced scripting or optimisation methods sold separately at similarly elevated cost, require significant manual intervention, often offer limited flexibility, and are not optimised for cryogenic applications, making them prohibitive for small research groups. To overcome these limitations and support precise cryogenic device calibration, a 65 nm chip containing some stand-alone transistor structures, previously fabricated by the Cryo-CMOS group at MCCI, was measured at both room temperature and 4 K to capture the I–V characteristics required for calibration. Building on these measurements, the first key contribution of this thesis is IceMOS, a Python-based tool developed to automate several recursive calibration tasks. IceMOS uses experimental I–V curves measured at both temperatures into the BSIM4 device model and generates all required spice simulation netlists. The user then iteratively adjusts selected model parameters, observing simulated curves alongside laboratory measurements until the error is minimised. IceMOS also automates device-model extraction, formats measurement data into CSV for streamlined comparison, and exports calibrated parameters fully compatible with standard PDKs. By significantly reducing manual effort and computational time, IceMOS enables faster and more efficient calibration of cryogenic CMOS models, achieving median errors below 5\% for PMOS and NMOS devices operating in the strong-inversion region. Following standard calibration methodologies reported in the literature, where comprehensive characterisation across all device bins is required, a second 65 nm chip was subsequently designed and fabricated. This chip includes a complete set of transistor bins, along with a simple amplifier structure. As a second contribution, this thesis also presents testing results obtained from this new chip under cryogenic conditions. In both the literature and the IceMOS calibration flow, parameters such as threshold voltage and slope factor are often initially estimated, a practice that can introduce significant inaccuracies. A third contribution of this thesis is the application of the S-EKV model combined with a Z-score filtering technique to systematically obtain precise initial values for these critical parameters, thus eliminating the reliance on estimations. This filtering approach was particularly important when addressing measurement-quality challenges at 4 K, where noise in the subthreshold and transition regions can significantly impact parameter extraction. The Z-score algorithm effectively removed outliers from the raw measurement data, reducing extraction errors from above 10\% to under 2\%, and significantly improving the reliability of the calibrated models. By first obtaining accurate S-EKV parameters and then using them as starting points for BSIM4 calibration, the methodology achieved excellent results with both the commercial 65 nm node and the open-source Sky130 130 nm technology, the latter leveraging publicly available datasets. The practical effectiveness of both IceMOS and the Z-score filtering technique has been demonstrated through experimental validation and reported in peer-reviewed IEEE conference publications. Overall, this thesis represents foundational work, presenting robust methodologies, thorough documentation, and practical tools designed to significantly advance research and development in Cryo-CMOS technology.Item Design and integration techniques for compact, wideband and multi-functional baluns(University College Cork, 2024) Steele, Joe; Psychogiou, Dimitra; Science Foundation Ireland; SkyWorks Inc.The prevalence of RF components in 5G communications and the ‘Internet of Things’ (IoT), has resulted in the integration of electronic devices into all facets of life. This has driven the need for RF components with smaller footprints, lower power consumption and less loss, while being more versatile across the ever-more crowded frequency spectrum. RF baluns, being a key front-end passive component in RF communications must meet these requirements, while effectively providing an impedance transformation and converting balanced-to-unbalanced RF signals. However, current balun technologies are typically large in size, lossy, or narrowband, calling for the development new design techniques and integration methods that will overcome these modern challenges. On the basis of the aforementioned limitations, this dissertation investigates RF and EM design methodologies for the realization of broadside-coupled line baluns with highly-miniaturized footprint, ultra-wide bandwidth (BW), and the added RF capability of bandpass filtering. Furthermore, the thesis outlines the performance capabilities and limitations of commercially available streamlined multilayer PCB processes and emerging additive manufacturing technologies using inkjet printing. Specifically, this thesis lays the foundations for; i) maximizing the fractional bandwidth (FBW) of Marchand baluns (MBs) and capacitively-loaded variants, dependant on process constraints, using broadside coupled transmission lines and perforated ground planes, ii) enhanced footprint miniaturization using self-packaged integration concepts enabled by freeform inkjet printing, and iii) RF co-designed bandpass filtering, by utilising the resonant behaviour of the coupled-line sections and exploiting the inkjet printing process to tailor their coupling levels. Experimentally validated results that are presented in this work demonstrate i) FBWs up to 122%, larger than what is achievable with the most common balun implementations, ii) footprints below 0.005 λg2 , which are competitive with state-ofthe-art miniaturised balun designs, and iii) 3rd-order compact RF co-designed filtering baluns, with FBWs from 30-110%, and footprints at least 6x smaller than what is currently presented in the literature.Item A system for efficient environmental monitoring and detection of forest fires through LoRa IoT and Artificial Intelligence(University College Cork, 2025) Uremek, Ipek; Popovici, Emanuel; Leahy, PaulForest fires are among the most pressing global challenges, causing devastating environmental, economic, and social consequences. With climate change intensifying the frequency and intensity of wildfires, the need for innovative, efficient, and scalable early warning systems has become more critical than ever. This thesis presents a versatile, cost-effective software and hardware platform for forest fire detection and prevention, leveraging advancements in low-power IoT protocols and Artificial Intelligence (AI) to address this global issue. The proposed system integrates a network of low-cost, LoRa-enabled sensors to monitor key environmental parameters such as temperature, humidity, and atmospheric pressure across vast forest ecosystems. LoRa (Long Range) technology provides an ideal communication solution due to its ability to transmit data over long distances with minimal energy consumption, ensuring the sustainability of the system in remote and resource-constrained areas. Central to this system is the collaborative interaction between a localized model and a global model. The localized model operates at the node level, performing anomaly detection under resource constraints. Meanwhile, the global model, represented by meteorological centers, enhances predictions by integrating high-quality environmental datasets. ARIMA is employed for time-series forecasting, analyzing environmental trends, and enabling better model interplay, while decision tree algorithms are utilized for fire risk assessment, providing critical insights into potential fire occurrences. The system’s protocol integrates these components efficiently. LoRa-enabled sensors collect real-time environmental data, including temperature, humidity, and air pressure, and transmit it via LoRaWAN. At the node level, localized models perform initial anomaly detection, generating alerts and probability scores. These, along with raw sensor data, are sent to a central system, where a collaborative framework refines the analysis for improved fire prediction. To enhance reliability, the system incorporates a feedback mechanism that compares sensor readings with meteorological data from the Irish Meteorological Service, mitigating issues caused by noisy or incomplete data. This hybrid approach, combining low-power IoT protocols with machine learning enhances wildfire detection accuracy while maintaining energy efficiency and reducing operational costs.Item Hypoxic-ischemic encephalopathy grading using novel EEG signal processing and machine learning techniques(University College Cork, 2025) Twomey, Leah; Popovici, Emanuel; Temko, Andriy; Qualcomm\IEEEPARstart{H}{ypoxic-Ischemic Encephalopathy} (HIE) is a critical brain injury in newborns resulting from oxygen or blood supply deprivation during birth. Traditional diagnosis methods for HIE require specialized expertise from a trained medical professional, and immediate intervention of treatment is required within the first 6 hours post primary hypoxic-ischemic insult. Electroencephalogram (EEG) is the gold standard for analysing brain pathologies for neonates and large windows of this complex signal must be assessed for accurate HIE diagnosis. Using an openly available dataset, this thesis proposes an automated approach for HIE grading, leveraging signal processing techniques and Artificial Intelligence (AI) to provide accurate and timely assessments, with an implementation enabling edge-device deployment for real-world clinical utility. Representing the 1 hour epoch of the EEG signal in the amplitude and frequency domain through Mel Spectorgram representation, the HIE grading task becomes an image recognition problem, where Convolutional Neural Networks have shown high accuracy and efficiency. An initial test accuracy of 84.85\% is achieved. Further enhancement of the signals rhythmic behaviour is necessary to increase the grading potential of the signal, thus the FM/AM sonification algorithm was implemented, transforming the EEG signal into an amplitude and frequency modulated audio signal. This pre-processing is designed to enhance the background rhythmic pattern of the signal, an essential feature used by the clinician for visual inspection. The two dimensional (2D) CNN is designed as a regressor to map the input image to a value on the HIE grading scale to enhance the model's ability to leverage the monotonic relationship between the grades. An optimised rounding function is implemented to define the final clinical grade as a novel post-processing technique. An overall test accuracy of 89.97\% based on a rigourous nested cross-validation evaluation framework is achieved, surpassing the current state-of-the-art by 6\%. The robustness and generalisability of the proposed method is obvious from the weighted F1-score of 0.8985 and a Kappa score of 0.8219. While enhancing the accuracy and speed of HIE diagnosis with this AI-driven approach, the goal is to make it readily available at the point of care. An inference time of 62.7 milliseconds is achieved for the quantized model on the Snapdragon processor highlighting its suitability for real-time, on-device HIE grading.Item Interactive EEG visualisation in Virtual Reality: design and implementation(University College Cork, 2025) Creed, Adam; Popovici, Emanuel; Factor, Andreea; Murphy, David; QualcommElectroencephalography (EEG) has long been a cornerstone in both neuroscience research and clinical diagnostics, providing invaluable insights into the electrical activity of the brain. EEG is the gold standard to detect and analyse brain pathologies such as seizures. An early detection and diagnosis of seizures is key in an effective treatment and decrease in morbidity and mortality. However, neonatal seizures are difficult to detect through clinical signs and even through EEG. The standard method of analysis of neonatal EEG is through visualisation on conventional two dimensional (2D) displays. Despite its importance, interpreting the complex and often nuanced signals produced by EEG remains a significant challenge, particularly when viewed on conventional 2D displays. This limitation can be particularly evident for clinicians who may lack extensive experience in EEG analysis. As the demand for more accurate and accessible diagnostic tools grows, the need for more intuitive methods of data visualisation becomes increasingly apparent. This thesis addresses these challenges by presenting the design, development, and comprehensive evaluation of an innovative neonatal EEG visualisation platform within a Virtual Reality (VR) environment, developed using Unity. The platform reimagines how EEG data can be presented, leveraging the immersive capabilities of VR to offer a fully three-dimensional (3D) space where users can interact with and explore brain activity data in real time. By moving beyond the constraints of 2D screens, this approach provides a more natural, immersive, and intuitive framework for both novice and expert clinicians alike. Recently, another method of analysis through sonification of EEG was proposed to speed up analysis of EEG. This allows users to hear the brain’s electrical activity as a dynamic auditory experience. This novel sonification technique not only provides an additional sensory modality for interpreting EEG data when used complementary with visualisation but also enhances the users’ ability to detect patterns and anomalies in the brain’s electrical signals that may be overlooked visually. Therefore, it was added to this platform. Furthermore, the platform integrates an AI-driven seizure detection algorithm. This feature maps detected seizure events onto a 3D brain model, offering clinicians a more spatially accurate representation of potential seizure zones. The ability to visualise these detections in a 3D context is expected to improve clinical decision-making. To validate the effectiveness of this VR-based EEG platform, a series of user evaluations was conducted. Participants, including both clinicians with EEG knowledge and those without prior experience, interacted with the system and provided feedback on its usability and functionality. The results were overwhelmingly positive, with the platform achieving a high System Usability Scale (SUS) score of 83, indicating strong user satisfaction. Participants also completed the NASA Task Load Index (NASA-TLX), with an average score of 36.65, reflecting a low perceived cognitive workload during the interaction. This suggests that the platform not only offers a rich and engaging user experience but also minimises the mental effort required to interpret complex EEG data. In addition to its intuitive user interface, the platform was rigorously tested across a range of hardware configurations, from high-end VR systems to more affordable, lowerend devices. It performed consistently well, demonstrating its potential for widespread adoption in diverse clinical environments, from cutting-edge research facilities to more resource-constrained healthcare settings.