QoS-Aware Decision-Making Framework for Service Allocation in IoT Architectures
The increasing complexity of Internet of Things (IoT) environments requires adaptive mechanisms for efficient service allocation across distributed infrastructures. Existing approaches are often focused on specific optimization algorithms or isolated Quality of Service (QoS) parameters, lacking a unified framework for decision-making across IoT, Fog/Edge, and Cloud layers. This paper proposes a modular QoS-aware decision-making framework that integrates QoS profiles, key performance indicators (KPIs), utility functions, and multi-criteria decision-making mechanisms to support both static and dynamic service allocation. The framework considers service execution latency, energy consumption, network throughput, and network coverage as decision criteria and enables adaptive balancing of conflicting QoS requirements. Its applicability is demonstrated through a smart transportation use case involving multiple service allocation scenarios across CRU, Fog, and Cloud infrastructures. The results confirm that different allocation strategies exhibit distinct QoS trade-offs and demonstrate the suitability of the proposed framework for adaptive and resource-efficient service allocation in layered IoT architectures.