Advanced trust evaluation and estimation solutions for the Internet of Things
View/
Keywords
Cybersecurity ; Cyber security ; Trust management ; IoT ; Trust score prediction ; Trust score calculation ; Transfer learning ; Few-shot learning ; Device onboarding ; Real-time predictionAbstract
The rapid proliferation of the Internet of Things (IoT) across critical infrastructure and various aspects of daily life, coupled with the growing sophistication and dynamic nature of cyber threats, highlights the urgent need for effective trust assessment within IoT ecosystems. This thesis begins by outlining the fundamental challenges associated with establishing and maintaining trust among IoT ecosystem entities. To address these challenges, it introduces a set of comprehensive trust evaluation, prediction, and management solutions.
The current research defines novel trust evaluation and prediction methodologies and identifies key factors that influence trust in IoT devices, particularly those providing services within the ecosystem. Furthermore, the thesis investigates adversarial strategies that aim to undermine trust assessment mechanisms—either directly, by corrupting the evaluation process, or indirectly, by manipulating predictive models.
To counter these threats, this dissertation integrates mathematical methods and statistical models in conjunction a suite of advanced technologies, including Few Shot Learning (FSL), Transfer Learning (TL), Machine Learning (ML), Intrusion Detection System (IDS), Distributed Ledger Technology (DLT), Physical Unclonable Function (PUF), and Trusted Execution Environment (TEE). This approach serves a dual purpose: (a) to reliably evaluate and forecast device trustworthiness in IoT environments, and (b) to detect and mitigate trust management-related attacks, as well as broader threats—such as Distributed Denial of Service (DDoS) attacks—targeting the components of the trust management system.
Finally, this thesis presents a suite of trust evaluation and prediction mechanisms specifically designed to meet the unique requirements of IoT ecosystems. It also provides a comprehensive assessment of the proposed methods in terms of security effectiveness and system performance.
Department
Σχολή Τεχνολογιών Πληροφορικής και Επικοινωνιών. Τμήμα Ψηφιακών ΣυστημάτωνNumber of pages
151Language
EnglishCollections
Except where otherwise noted, this item's license is described as
Αναφορά Δημιουργού-Μη Εμπορική Χρήση-Όχι Παράγωγα Έργα 3.0 Ελλάδα
Αναφορά Δημιουργού-Μη Εμπορική Χρήση-Όχι Παράγωγα Έργα 3.0 Ελλάδα