Ιnternet of Things (IoT) as a concept, refers to interconnected heterogeneous devices with different specifications and performance capabilities, which operate in resource-constrained environments. To guarantee the smooth behaviour of such IoT networks, we seek ways of constantly monitoring and detecting failures and malfunctions of the various modules as Quality of Service/Quality of Experience (QoS/QoE) services, which can be embedded as services into IoT platforms.
This thesis will evaluate the current state of the art in the domain of automatic and explainable QoS/QoE using statistical and AI/ML-based algorithms, propose a design, and implement a method for the automatic assessment of IoT services quality. As a proof of concept, the proposed approach will be integrated and tested in the research-oriented IoT platform of SKEL (SYNAISTHISI).