A5068029855- IoT and Edge/Fog Computing
- Software-Defined Networks and 5G
- Cloud Computing and Resource Management
- Caching and Content Delivery
- Software System Performance and Reliability
- Blockchain Technology Applications and Security
- Smart Grid Security and Resilience
- Advanced Memory and Neural Computing
- Cloud Data Security Solutions
- Mobile Ad Hoc Networks
- Distributed and Parallel Computing Systems
- Advanced Computing and Algorithms
- Cryptography and Data Security
- Robotics and Automated Systems
- Smart Cities and Technologies
- Advanced Optical Network Technologies
- Wireless Body Area Networks
- Scientific Computing and Data Management
- Digital Transformation in Industry
- IoT Networks and Protocols
- Access Control and Trust
- Music and Audio Processing
- Modular Robots and Swarm Intelligence
- Network Traffic and Congestion Control
- Complex Network Analysis Techniques
National Technical University of Athens
2018-2025
Institute of Communication and Computer Systems
2024-2025
Lancaster University
2022
With the advent of Internet Things (IoT), centralized cloud computing service delivery paradigm has been gradually transformed into a continuum that includes edge and fog heterogeneous IoT devices with varying power capabilities [...]
Network slicing is set out to address crucial needs of 5G, including support for multi-service provisioning. Focusing on resource and performance isolation, as well security concerns associated with multi-tenancy, existing management orchestration (MANO) frameworks typically offer network slices in the form isolated bundles computing, storage, resources across infrastructure, edge. However, slice instantiation its prevailing raises significant related utilization, hindering potential...
With the advent of 5G verticals and Internet Things paradigm, Edge Computing has emerged as most dominant service delivery architecture, placing augmented computing resources in proximity end users. The resource orchestration edge clouds relies on concept network slicing, which provides logically isolated resources. However, though there is significant progress automation within a single cloud or datacenter, multi-domain infrastructure multi-administrative domain still an open challenge....
With the rapid advancement in next-generation Internet of Things (IoT) and ever-expanding virtual representation devices, novel techniques definitions are key to accommodate emerging paradigms transform our world. In quest achieving coveted convergence IoT technologies with Cloud-Edge computing continuum seamless interaction devices hyper-distributed applications, we enhance concept Virtual Objects (VOs) for virtualization wake recent works. We introduce a software stack, called VOStack,...
The evolution of 5G network service ecosystems raises the need for discovery services among multiple (edge) clouds with potentially heterogeneous resources and features. Besides deployment, such a architecture could provide means cross-service communication (CSC), that is, consumption (NS) by another NS, empowering capabilities latter. This article presents distributed framework marketplaces (NSMs) in edge computing. We exemplify operation new mechanism relies on cache-based protocol NSs...
The evolution of Internet Things (IoT) has sparked significant research interest in edge computing. Within this scope and given the ever-increasing number IoT mobile devices, computation offloading is emerging as a cutting-edge area with enormous potential practical applications.In respect, we present architecture design experimental evaluation an orchestration framework for smart from or devices to cloud servers. proposed platform, namely COSMOS, includes control-plane components workload...
In the era of 5G networks, demand for high quality service provisioning is growing extremely fast. The enabling Network Function Virtualization and Slicing in scope network aims to meet strict requirements various business cases. Alongside, complexity deployment such services becomes also higher, regarding differences between infrastructure capabilities plethora individual requirements. This fact makes selection appropriate slice a complex, but also, major process, as optimization leads...
A federation of heterogeneous testbeds, which provides a wide range services, attracts many experimenters from academia and industry to evaluate novel future Internet architectures network protocols. The candidate experimenter reserves the appropriate testbeds’ resources based on various diverse criteria. Since several testbeds offer similar resources, trust mechanism between users providers will facilitate proper selection testbeds. This paper proposes fuzzy reputation-based framework that...
The consolidation of 5G technologies makes the Industry 4.0 establishment more feasible than ever. Network Function Virtualization (NFV) and Edge Computing aim at managing multi-vendor industrial network infrastructure guaranteeing high-performance requirements processes. Furthermore, production workflows require collaboration between robots, actuators sensors in factory floor. In this direction, Slicing allows deployment complex end-to-end services, enabling applications different vendors...
The Internet of Things (IoT) paradigm, with a grand variety devices in large numbers, commonly combined resource and power limitations dictates new challenges constraints. emerging Edge Computing paradigm arises promises to address these issues by placing cloud-type resources, closer IoT devices. In this context, network slicing is common practice at the edge. A slice groups logically isolated computing resources. Under 5G umbrella, multi-administrative domain services have emerged as...
Maximizing the performance of modern applications requires timely resource management virtualized resources. However, proactively deploying resources for meeting specific application requirements subject to a dynamic workload profile incoming requests is extremely challenging. To this end, fundamental problems task scheduling and autoscaling must be jointly addressed. This paper presents scalable architecture compatible with decentralized nature Kubernetes [1], solve both. Exploiting...
In this study, we approached the Hit Song Prediction problem, which aims to predict songs will become Billboard hits. We gathered a dataset of nearly 18500 hit and non-hit extracted their audio features using Spotify Web API. test four machine-learning models on our dataset. were able success song with approximately 86\% accuracy. The most succesful algorithms Random Forest Support Vector Machine.
Autoscaling resources in a power-efficient way is essential to enable Green Computing resource management solutions. The development of dynamic provisioning techniques could lead the minimization power consumption and simultaneously guarantee high quality service (QoS) inline with workload demand. In this work, we introduce AHP4HPA, an autoscaling framework for Kubernetes Clusters, which aligned architecture state-of-the-art practices. We define profiles, namely mapping between QoS computing...
In the dawn of 6G era, evolving service ecosystems raise need for cross-service interactions. Existing resource/service orchestration frameworks are oblivious to such interactions, since they tend orchestrate services independent each other. this respect, we stress on optimized crossservice communication, and more specifically, cross-slice communication (CSC). To end, present a CSC platform, namely MESON, which fulfills all main requirements, as discovery CSC-enabled services, selection most...
The evolution of modern applications and the adoption 5G network slicing architecture require minimizing communication delays. Virtual Network Functions (VNFs) offer low latency services to remote users. In context Edge Computing, these virtualized are hosted in Clouds, which, however, have limited resources. Embedding Slice requests from users is a challenging problem, demands distributed solutions presence resource delay constraints Network. This paper introduces delay-aware service...