Mid Sweden University

In the realm of intelligent transportation systems, vehicle-to-infrastructure technology is a cornerstone for improving road safety. However, it also highlights the need for effective task offloading strategies across the edge-fog-cloud continuum. This paper presents a testbed for evaluating task allocation algorithms designed to optimize resource utilization while meeting real-time requirements. Therefore, a simulated V2I environment testbed has been developed to demonstrate and evaluate the potential of different algorithms to improve computational efficiency across the continuum. Our results therefore include an implemented testbed system to emulate, test, and evaluate different algorithms, laying a foundational step towards more sophisticated offloading strategies in the future. This will enable future research and further advances in optimizing communication and computation across the edge-fog-cloud continuum.

In the rapidly evolving cybersecurity landscape, detecting and preventing network attacks has become crucial within the industrial sector. This study aims to explore the potential of intrusion detection by employing deep learning within edge computing, especially for the Industrial Internet of Things. Specifically, TinyML converted CNN, LSTM, Transformer-LSTM, and GCN models on the UNSW-NB15 dataset. A comprehensive dataset analysis gained insights into the nature of attack behavior data. Subsequently, a comparative analysis in an edge computing setup using Raspberry Pi units revealed that the GCN model, with its accuracy of 97.5%, was the best suited of the compared models for this application. However, the study also explored variables like time consumption, where the CNN model was the fastest out of the compared models. This research also highlights the need for continued exploration, especially in addressing dataset imbalances and enhancing model generalizability. By recognizing each model's strengths and areas of improvement, this research serves as a step toward bolstering digital safety and security in an increasingly interconnected industrial world.

Vi har tagit fram en modern läskurs i seminarieform, där studenterna själva sätter upp ett självstudieprojekt genom att dra nytta av MOOC-kursmaterial. Vår kurs avser att ge civilingenjörsstudenter erfarenhet av denna typ av online-kursmaterial och på så sätt förbereda dem på ett framtida livslångt lärande. Kursen ger även fördjupningsmöjligheter inom många fler områden än lärosätet kan erbjuda konventionella kurser. Vidare syftar kursen till att förmedla både teoretiska kunskaper och praktiska färdigheter som förberedelse inför utbildningens avslutande självständiga arbete, samt att möjliggöra specialisering som start på en framtida karriärväg. Vi har med små lärarinsatser skapat detta kursupplägg, genom att tillämpa studentcentrerat lärande och flipped classroom-tekniker. Studenterna kompletterar ofta MOOC-kurserna med djupare teoriuppgifter, positionerar ämnet vetenskapligt, och reflekterar över sitt lärande i en slutrapport. Vi examinerar hur väl de har planerat och genomfört sina självstudier. Kursen har erbjudits för fyra årskullar. Genomströmning har varit 93% och den har fått goda omdömen i kursutvärderingar. Eftersom studenterna kan välja mellan ett stort antal MOOC kurser som ständigt utvecklas, och studenterna brukar välja aktuella ämnen, så håller kursen sig själv uppdaterad inför framtida trender. Vi lärare behöver endast följa med på resan och får kompetensutveckling inom nya områden på köpet.

This article explores a solution using fog computing and Internet of Things technologies to address some of the drawbacks with cloud computing. We show the potential in the combination of scalable distributed systems technologies, for enabling a distributed cluster using typical lightweight IoT devices. Answering research questions related to the expected overhead, as well as how work and chunk size affect the computational time and scalability of the system. All in order to evaluate the performance and potential of this technology for Industry 5.0 and future industrial IoT applications. Our results show the great potential of the approach, but we also identify necessary improvements that needs to be made in order to achieve performance superiority over the traditional cloud approaches and for the technology to truly proliferate.

I dagens samhälle blir det allt viktigare att fortbilda sig under hela sitt yrkesverksamma liv. För att möta efterfrågan på det livslånga lärandet har Mittuniversitetet utvecklat och genomfört ett antal kurser som riktar sig mot yrkesverksamma ingenjörer. Detta arbete presenterar våra erfarenheter av att ge dessa kurser, med en tyngdpunkt på studenternas upplevelser. Syftet med detta är att bygga upp en vetenskaplig bas för vad vi gör som är bra, men även vad som kan förbättras och förändras. Målsättningen är att göra dessa specialanpassade kurser riktade mot yrkesverksamma ingenjörer så givande och flexibla som möjligt. Våra initiala resultat visar bland annat att studenternas negativa upplevelser ofta var kopplade till antagningsförfarandet och det praktiska genomförandet av kurserna. Man hade svårigheter med att hitta hur man skulle registrera sig på kursen och att tidsramen för registrering kunde vara ett problem. Läroplattformen uppfattades som svår att överblicka och det förekom även viss otydlighet gällande var undervisningen skulle äga rum. Den positiva responsen i utvärderingarna gällde främst det faktiska kursinnehållet, då man ansåg att uppgifter och kursmaterial var givande. Vidare uppskattades kursupplägget, att man kunde kombinera studierna med arbete. Framledes kommer vi att fortsätta med dessa utvärderingar i takt med att kurserna ges, och därefter anpassa vårt mottagande och kommunikationen med studenterna. Även kursupplägget ses över kontinuerligt via den återkoppling vi mottar. 

The area of computer and electrical engineering is under constant evolution which leads to lifelong learning being an important aspect for being a long term successful working professional. This work presents an education model for creating customized and flexible courses at advanced level for these working professionals. We have approached this work from the educator's point of view and the focus will be on the teaching model and our results from implementing the model during the last two years. Including how we created these need and trend-based education offerings, the course execution inspired with micro-learning and flipped-classroom pedagogics, and our work with creating win-win possibilities within the courses for the working professionals and the companies they work for. Finally, we will present our experiences and lessons learnt, ending with a plan for our upcoming courses and our refined model in our ongoing future work.

Modern IT systems have been transitioning from traditional on-premises solutions to a dynamic mixture of on-premises and off-premises solutions. This transition has also included a trend to run more systems on software-defined resources. The ease of setting up new software-defined servers and systems has led to an increase in IT system complexity as well as the amount of log data generated. Automatic log analysis has become a subject of interest because of the problems with manual log analysis in case of intrusion detection and root-cause analysis. Therefore, this paper proposes and tests a sequence based anomaly detection method. The work has been done in collaboration with the Swedish Social Insurance Agency's IT department. Real system log data with high privacy requirements and limited available information was generated for training and testing. The generated log data was produced with expected time regions of anomalous behavior. Our proposed anomaly detection model was then able to perform at a state-of-the-art level and could accurately detect certain error types. Showing the potential of the approach when applied directly to a real-world system.

This research presents a solution for collecting indoor environmental sensor values and how the gathered sensor values then could be used for green building certification and in turn also machine learning based smart building control. We have created and implemented a proof of concept system consisting of a sensor collecting device using off the shelf hardware to complement the existing sensor information from buildings, as well as a cloud system for persistently storing this data for later usage. We have measured and evaluated our implemented system for our envisioned scenarios. In which we could observe that our proof-of-concept could scale to handle almost four sensor value updates per second at maximum stress, as well as having a latency for uploading a sensor value from our sensor of about 130 ms. Finally, we present our future and ongoing work based on these results which outlines our work for smart building control, green building certification, and the energy signature of buildings. 

The evolution from local toward virtualized data storage, computing, applications and services – in the forms of Internet of Things (IoT), Everything as a Service (EaaS), and Cloud computing – has changed the way of delivering digital services for consumers and businesses. These technologies have brought clear benefits over traditional systems, such as easy management, universal availability, and decreased hardware requirements for end-user devices. The main challenges of the first-generation IoT services are related to communication latency due to high physical and logical distance between end nodes and server resources and vulnerability for network problems along the long routes. Thanks to the unveiling of the fifth-generation wireless technology (5G) for cellular networks, the last-mile connection performance – communication latency in particular – is taking a huge leap and therefore introducing new possibilities for industrial applications. Edge computing is a key technology to unleash the full potential of the arising industrial wireless communication, since it enables deploying computational tasks to computing nodes near the end devices and therefore opens novel business opportunities around real-time cloud services. In this chapter, we introduce the current state of the art and discuss different challenges the edge computing systems are facing particularly in the Industrial IoT (IIoT) domain, as well as present potential solutions for the identified challenges.

Modern enterprise IT systems generate large amounts of log data to record system state, potential errors, and performance metrics. Manual analysis of log data is becoming more difficult as these systems become more complex. Therefore, machine learning based anomaly detection of system logs is a vital component for the future of system management. Existing log anomaly detection models commonly rely on learning the general normal behavior of the target systems to accurately detect anomalies. They are however limited by the often sparse existing system knowledge. Therefore, this paper proposes a general anomaly detection method which requires little or no knowledge of the target system. This is done by assuming there are semantic similarities in different systems’ log data. Labeled log data from other systems can then be used for training the anomaly detection model. The model uses self-attention transformers and ensemble learning techniques to learn the semantic representation of normal and abnormal log messages. The proposed method achieves a performance comparable to other log anomaly detection methods while requiring little knowledge of the target system.