ISEP - DM - Engenharia de Sistemas Computacionais Críticos
Permanent URI for this collection
Browse
Recent Submissions
- Performance monitoring of real-time applications in RISC-V platformsPublication . Soares, Nuno Filipe Pessoa; Carvalho, Tiago Diogo Ribeiro deReal-time systems are an important field of research, specially when considering that these systems interact with the real-world and their tasks have direct impact in human lives and society. It is imperative that the traceability and performance of these systems can be ensured. RISC-V is an open-source instruction set architecture, gaining interest from both industry and academia, specially in the context of real-time systems development due to its versatility and customization capabilities. The performance analysis of RISC-V systems emerges as a critical aspect, with the need to evaluate metrics, such as system efficiency and scalability, in order to deal with increasing workloads and tasks complexity. There is a lack of support for performance analysis, specially in real-time operating systems for RISC-V architectures, usually not providing full support to the latest performance monitoring related specifications, or being tangled to a specific operating system. This thesis provides a comprehensive study on performance analysis approaches on RISCV systems, exploring tools and solutions for performance monitoring of applications, that provide access to the system performance counters. As a solution, this work proposes an API capable of retrieving performance metrics from the hardware performance counters, accessible via code instrumentation over a target application. The API proposed works as a proof-of-concept for the development of more sophisticated tool capable of facilitating the retrieval and configuration of performance metrics from a RISC-V system. To access the API performance, some introductory tests have been performed to showcase that the API is capable of interacting and managing the hardware performance counters present on the system. Additionally, the overhead generated from the API was also analysed briefly, showcasing that the API has limited impact on the overall system’s performance when retrieving metrics from the hardware performance counters.
- Real-time quality index estimation for redundant sampled values streams in digital substationsPublication . Oriente, Everton Matheus; Barros, António Manuel de SousaProtection devices are essential elements in the electric power grid, as they safeguard equipment and ensure the stability and reliability of the network by detecting faults and preventing damage through timely isolation of affected areas. With technological advancements, electromechanical protection relays have evolved into Intelligent Electronic Devices (IEDs), essentially protection relays with built-in communication capabilities. These modern relays not only protect equipment but also exchange critical information with other devices. However, the variety of events and data exchanged between power grid components is substantial. In this respect, communication protocols present both advantages and challenges. On one hand, they enable the signalling of diverse events and the transmission of multiple pieces of information over a single medium. On the other hand, a protocol is only effective if all components within the system comply with it. The introduction of the IEC-61850 standard has established a unified communication across the energy sector, promotes interoperability between manufacturers, offering significant advantages over other standards. Real-time data exchange is facilitated by specific protocols within the IEC 61850 framework, which specifically address the communication requirements of IEDs, ensuring enhanced data quality and availability. As IEDs play a crucial role in substations, with many of their functions being both critical and time-sensitive, any developments incorporated into these devices can significantly impact the entire system. An IED analyses the stream of Sampled Values (SVs) sent by a Merging Unit to detect faults and implement protective measures. The Merging Unit sends SVs over two independent channels: the primary and the secondary (redundant) channels. When the primary channel fails, the secondary channel takes over, ensuring the information that the primary channel could not transmit is conveyed. As such, the current decision-making process for switching channels is based solely on the occurrence of a total channel failure. This thesis aims to enhance IED functionality by adding intelligence to this channel selection method. It proposes ways to evaluate, quantify, and qualify the received information to determine the most reliable signal source, whether from Channel 1 (primary) or Channel 2 (secondary). This approach ensures that protection relays receive better information, enabling protection algorithms to act correctly and promptly within their operating parameters.
- Variable pitch system for UAV proprotorsPublication . Rijo, Carlos André Pinto Ramos; Severino, Ricardo Augusto Rodrigues da SilvaUnmanned Aerial Vehicles (UAVs) have gained significant attention in diverse applications, prompting a surge in research and development. Current Vertical Take-Off and Landing (VTOL) UAVs rely on fixed-pitch proprotors which makes them face limitations in performance and maneuverability. This thesis addresses these challenges by proposing the development of a stand-alone Variable Pitch Proprotor (VPP) system capable of real-time pitch adjustments during different flight phases. The research delves into the fundamentals of VTOL flight performance, variable proprotor impact, control subsystems, power management, firmware strategies and short-range wireless communication protocols. Studying and analyzing existing solutions and technologies in these domains is conducted. The subsequent stages involve designing the subsystem architecture, implementing it on a mechanical prototype, and evaluating the system’s performance and limitations in various settings. This Thesis will contribute to enhance UAVs stability, maneuverability, and energy efficiency, towards to the advancement of Urban Air Mobility (UAM).
- Formal analysis of an authority transfer protocol for UAVsPublication . Rodrigues, Ricardo Alexandre Almeida; Pereira, David Miguel RamalhoThe use of Unmanned Aerial Vehicles (UAVs) in a variety of applications has resulted in a growing demand for Beyond Visual Line of Sight (BVLOS) operations to cover long distances. This thesis focuses on an initial analysis of a UAV communication protocol designed for transferring control authority between ground controls within a common range, addressing challenges and evaluating its efficacy across different operational phases. The research encompasses start-up, mission execution, and handover phases, assessing the protocol’s interactions with ground-based systems for reliability, efficiency, and adaptability. A comprehensive overview of model checking is provided, emphasizing transition systems, formalisms, and tools for system correctness verification. The research employs model checking to formally specify and model the UAV communication protocol, establishing evaluation criteria through the use of LTL and CTL properties. The derived requirements serve as a foundational framework for future iterations, ensuring the protocol’s robustness and addressing security strategies in the event of connection loss. The holistic approach contributes to a comprehensive understanding of the protocol’s functionality, aiming to enhance the security authorization handover procedure and promote public and regulatory acceptance of BVLOS operations with UAVs.
- Intelligent edge orchestration of real-time applicationsPublication . Calai, Rafael; Barros, António Manuel de SousaWith the rapid development of embedded systems, new communication technologies, and distributed systems, the complexity of critical system development has increased. In the past, these systems typically only performed the function of controlling a specific process, without the need for communication or being components of a complex system centralizing data processing for decision-making. Nowadays, these critical systems have more tasks and are much more integrated, aiming to provide new functionalities to users, reduce costs, and enhance the security and availability of essential services. This scenario has brought new challenges to the development of such systems, including increased computational needs, new communication methods, ensuring cybersecurity, and meeting real-time requirements in distributed systems. The central theme of the thesis will focus on the challenge associated with ensuring real-time requirements in distributed systems.To address this issue, we propose a scheduling algorithm for Docker Swarm, built on top of RT-Linux. This algorithm is designed to ensure that each process meets its temporal constraints by prioritising allocation according to their respective priorities and the resources available.
- Integrated monitoring for cyber-security in residential scenariosPublication . Carneiro, Mário Miguel Silva de Sá; Ferreira, Luís Miguel Moreira LinoThe increasing number of Internet of Things devices and the increasing adoption of smart homes have led to an increase in risk related to cybersecurity. The purpose of this dissertation is to examine these limitations and propose innovative solutions for anomaly detection using machine learning (ML) methods. The objectives and motivations for this work, which focuses on improving home network security, are explained in Chapter 1, which also provides a review of smart homes and their connections with cybersecurity issues. In Chapter 2, "State of the Art," the rise of IoT in everyday use and related safety concerns are addressed. In addition, it covers over basic concepts like machine learning strategies and the way these interact with intrusion detection systems (IDS). In order to mitigate increasing threats it considers that integrating ML with cybersecurity in IoT systems is important. The machine learning techniques selected for this project are presented in Chapter 3, with a focus on creating a reliable anomaly detection pipeline.Comprehensive data pre-processing, including cleaning, merging, normalization, and analysis, ensures sure the data is suitable for model training. In Chapter 4, training individual and ensemble models will be addressed along with an analysis of performance metrics in scenarios using binary and multi-class classification.Z-Score normalization is one strategy that is frequently used to handle unbalanced datasets. It has been demonstrated to perform better than Min-Max, especially when applied to the UNSW-NB15 dataset. The implementation of APIs using Streamlit for real-time visualisation and FastAPI for backend integration with ML models will be discussed in Chapter 5.This combination enables the ability to anticipate cyberattacks and visually represent anomalies in an effective way. The testbed built to automate cyberattacks and extract important features for model training will be discussed in detail in Chapter 6. The results of the evaluations, that compare the performance of the individual models and the ensemble, are presented in Chapter 7. The ensemble performed better than expected, especially when it came to identifying anomalies in multi-class environments. It achieved this with high accuracy and a significant reduction in false positives and negatives. Chapter 8 ends with an overview of the project’s conclusions and contributions.The main findings emphasize the significance of selecting normalization strategies and the advantages of using ensemble models to improve attack detection.
- Deployment of ML Mechanisms for Cybersecurity in Resource-Constrained Embedded SystemsPublication . Vicente, Pedro Miguel Casal; Santos, Pedro Miguel Salgueiro dosThe increase of low security devices in the Internet is being exploited by hackers to compro mise data or use to use them as external agents to perform further attacks. As so, it is of crucial importance that networks posses a system that correctly assess the nature of incom ing and outgoing packets to protect the local network and the overall Internet connected systems. To achieve this, Machine Learning is being broadly used due to his early success. Nevertheless, these mechanisms are better inserted at the entry point of local networks, an embedded system which has limited resources to train machine learning models and/or to perform inference tasks. Since Cybersecurity is a real-time problem, the embedded systems should perform these activities in a very restricted time interval. The time required to clas sify the packets depends on the overall system load, machine learning models complexity and desired accuracy. This thesis aims to assess the current support for ML in embedded systems, either through the interoperability of models or through their development in low level languages, and the relationship between the time required by different embedded sys tems, the different tools and models. This thesis explored one transpilation tool, m2cgen, two interoperability formats, PMML and Open Neural Network Exchange (ONNX) and one real time environment, ONNXRuntime, to deploy an already trained model in a device with limited resources. Results demonstrate that ONNXRuntime was the only machine learn ing tool with a perfect match regarding samples prediction’s classification from the original models. An analysis on the time required to execute this task revealed that ONNXRun time is faster than Scikit-Learn with the Isolation Forest (ISO), One Class Support Vector Machine (OCSVM) and Stochastic Gradient Descent One Class Support Vector Machine (SGDOCSVM) models and slower with the Local Outlier Factor (LOF) model.
- Towards the simulation of cooperative perception applications by leveraging distributed sensing infrastructuresPublication . Oliveira, Miguel Ferreira; Severino, Ricardo Augusto Rodrigues da SilvaWith the rapid development of Automated Vehicles (AV), the boundaries of their function alities are being pushed and new challenges are being imposed. In increasingly complex and dynamic environments, it is fundamental to rely on more powerful onboard sensors and usually AI. However, there are limitations to this approach. As AVs are increasingly being integrated in several industries, expectations regarding their cooperation ability is growing, and vehicle-centric approaches to sensing and reasoning, become hard to integrate. The proposed approach is to extend perception to the environment, i.e. outside of the vehicle, by making it smarter, via the deployment of wireless sensors and actuators. This will vastly improve the perception capabilities in dynamic and unpredictable scenarios and often in a cheaper way, relying mostly in the use of lower cost sensors and embedded devices, which rely on their scale deployment instead of centralized sensing abilities. Consequently, to support the development and deployment of such cooperation actions in a seamless way, we require the usage of co-simulation frameworks, that can encompass multiple perspectives of control and communications for the AVs, the wireless sensors and actuators and other actors in the environment. In this work, we rely on ROS2 and micro-ROS as the underlying technologies for integrating several simulation tools, to construct a framework, capable of supporting the development, test and validation of such smart, cooperative environments. This endeavor was undertaken by building upon an existing simulation framework known as AuNa. We extended its capabilities to facilitate the simulation of cooperative scenarios by incorporat ing external sensors placed within the environment rather than just relying on vehicle-based sensors. Moreover, we devised a cooperative perception approach within this framework, showcasing its substantial potential and effectiveness. This will enable the demonstration of multiple cooperation scenarios and also ease the deployment phase by relying on the same software architecture.
- Controlo e Monitorização do Ar InternoPublication . Silva, Jorge Manuel e; Barros, António Manuel de SousaMaintaining reliable and secure operation of an organization's servers is an arduous task that requires continuous monitoring and control of several factors. Controlling the operating environment of a server (e.g. temperature and humidity) is critical, as deviations from the equipment's nominal operating conditions can compromise the processes that depend on them, with a consequent negative impact on the organization. Idealizing and prototyping a final product is a laborious process, as its conditions vary from organization to organization. For each case, it is essential to think, idealize, plan and, finally, execute. This project's main objective is to monitor and control the environmental conditions present in a rack through sensors and actuators. The sampled data is integrated into an application that determines the actions that allow maintaining the best operating conditions and preventing/detecting the occurrence of accidents in order to minimize the possibility of loss of resources and critical data.
- Authority Handover Procedure and Safety Decision Strategy in Unmanned Aerial VehiclesPublication . Ferreira, Gleizielly Alves; Tovar, Eduardo Manuel MedicisOver recent years, Unmanned Aerial Vehicles (UAVs) applications have become popular in different areas, such as aerial image acquisition, agriculture, inspection and maintenance, mapping, and delivery services. Some of these services require the ability to fly UAVs Beyond Visual Line of Sight (BVLOS) to cover greater distances. Data provided by onboard instruments control the BVLOS operation. The flight controller is responsible for directing the drone flight by controlling the motor’s speed and gathering sensor data. The relevant information about the aircraft, such as position, altitude, speed, and direction of flight, are transmitted via a radio link that informs an operator or a Ground Control Station (GCS). In some drone architectures, there is also an extra computer known as a companion computer or mission computer. They are responsible for providing more intelligence to the flight controller by changing flight parameters. The tasks running on a companion computer can add the capacity to make intelligent decisions during autonomous flight or emergencies, for instance, when the drone loses the radio link with GCS. In addition, for complex drone operations in larger coverage areas, it is necessary to transfer wireless communication links from one access point to another without experiencing connectivity loss. This procedure is known as Handover, and there is much research in this area. Therefore, studies in this field are still needed to find better solutions to avoid failures and increase public and regulatory acceptance of BVLOS operations with UAVs. In this context, the thesis intends to address solutions to the security authorization handover procedure and addresses security strategies in case of a loss of connection.