Abstract
dc:descriptionThe increasing adoption of cloud, fog and edge computing has reshaped how modern systems are designed and operated, placing resource efficiency at the centre of distributed computing research. As computation and storage move closer to data sources and end users, infrastructures become more heterogeneous, constrained and dynamic, revealing the limits of static optimization and centralized control. This thesis examines resource usage optimization in the cloud–edge continuum by treating efficiency as a multi-factor optimization issue that is both a continuous and workload driven process, influenced by data characteristics, system topology and deployment decisions. The work is structured around a sequence of representative use cases that span multiple layers of the systems, including big data analysis, power consumption forecasting, smart city data distribution, edge storage systems and proactive application placement. Across these scenarios, the thesis investigates how resource consumption patterns and optimization opportunities emerge under realistic operational conditions. The results show a strong corelation between data volume, velocity, representation and resource demand, as well as clear trade offs between responsiveness, stability and efficiency when comparing streaming and batch processing models, centralized and peer-to-peer storage architectures and different orchestration strategies. A common finding is that optimization is often constrained less by algorithmic complexity than by how data and services are organized, accessed and positioned within the system. Lightweight predictive models, when carefully optimized, are proven able to capture short term workload dynamics on resource limited edge devices without introducing significant overhead. Similarly, changes in data representation, storage layout and placement strategies are demonstrated to substantially affect latency, network utilization, energy consumption, and quality of service, even when analytical models remain unchanged. Building on these observations, the thesis introduces adaptive optimization frameworks that treat placement and orchestration as ongoing decision making processes, contributing practical insights and methodologies for improving resource efficiency across the cloud–edge continuum.
Degree
thesis:*- Grantor dc:publisher
- Harokopio University
- Year dc:date
- 2026
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Psomakelis, Evangelos
- Ψωμακέλης, Ευάγγελος
Subjects
dc:subject × 16- Cloud–Edge συνεχές
- Βελτιστοποίηση πόρων
- AI-υποβοηθούμενη ενορχήστρωση
- Συστήματα υπολογιστικού νέφους
- Βελτιστοποίηση πολλαπλών στόχων
- Cloud–Edge Continuum
- Resource Optimization
- AI-Assisted Orchestration
- Cloud-Native & Adaptive Systems
- Multi-objective optimization
- Επιστήμη Ηλεκτρονικών Υπολογιστών και Πληροφορική
- Φυσικές Επιστήμες
- Πληροφοριακά συστήματα
- Computer and Information Sciences
- Natural Sciences
- Information Systems
Rights
- Language dc:language
- eng
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/61297