Balancing QoE and QoS in Behavior-enabled IoT

Modern Smart Cities and distributed systems face the challenge of balancing technical efficiency with user satisfaction. Emerging paradigms such as the Internet of Behavior (IoB), Internet of People (IoP), and Internet of Senses (IoS) recognize human behavior as a first-class element in software architecture; yet, they often lack formal mechanisms to resolve the inherent conflicts between Quality of Service (QoS) and Quality of Experience (QoE). To address this gap, this thesis proposes an engineering framework based on the Behavior-Enabled (BeT) paradigm, where human and system components establish bi-causal interactions to jointly manage QoE and QoS. Leveraging Digital Twin (DT) technology, BeT systems address the tri-dimensional uncertainty (Self, Context, and Human) by modeling and analyzing the behaviors of people, physical environments, and software infrastructures. This enables asymmetric dual-action strategies: autonomous system reconfiguration (strong causality) and non-coercive behavior-guiding recommendations (weak causality). This causal interplay forms a continuous negotiation loop in which system configurations adapt to human behavior to optimize QoE, while user behavior is influenced to enhance QoS. The contributions of this work include: the design of the BeT conceptual meta-model, introducing a tridimensional quality framework that captures the interdependencies among QoE, Business QoS, and Cyber QoS; the design of a multi-twin reference architecture coordinated by a centralized BeT Manager; the sequence diagrams to offer a dynamic interpretation of the components' interactions; and the definition of a modeling process for designers and practitioners. The proposed reference architecture was directly shaped by the analysis of two significant Smart City scenarios that served as its foundational design drivers: Renewable Energy Communities (REC), which balance community energy efficiency with user routines, and Multimodal Public Transport, focusing on vehicle crowding versus passenger comfort. The architectural convergence of QoE and Cyber QoS is primarily driven by a novel approach to behavioral observability, which captures the complex interplay between computational resources and user perception. This then operationalized through the AURA framework, which leverages these behavioral insights to synchronize Human and Cyber Twins, ensuring the dynamic adaptation of ICT resources to sustain human performance in critical contexts.