A5065495505- Cloud Computing and Resource Management
- Distributed and Parallel Computing Systems
- IoT and Edge/Fog Computing
- Parallel Computing and Optimization Techniques
- Distributed systems and fault tolerance
- Real-Time Systems Scheduling
- Age of Information Optimization
- Advanced Queuing Theory Analysis
- Scheduling and Optimization Algorithms
- Interconnection Networks and Systems
- Blockchain Technology Applications and Security
- Anomaly Detection Techniques and Applications
- Digital Transformation in Industry
- Graph Theory and Algorithms
- Electricity Theft Detection Techniques
- Imbalanced Data Classification Techniques
- Industrial Vision Systems and Defect Detection
- Context-Aware Activity Recognition Systems
- Scientific Computing and Data Management
- Retinal Imaging and Analysis
- Traffic Prediction and Management Techniques
- Brain Tumor Detection and Classification
- Simulation Techniques and Applications
- Advanced Control Systems Optimization
- Caching and Content Delivery
Aristotle University of Thessaloniki
2015-2024
University of Cyprus
2023-2024
University of East London
2022
Wayne State University
2022
University of Warwick
2022
Nottingham Trent University
2022
Universitat Politècnica de València
2022
Brunel University of London
2022
Cardiff University
2021
Edge Technologies (United States)
2021
The ever increasing popularity of cloud computing has relieved many consumers and businesses from the burden acquiring, maintaining monitoring expensive hardware software infrastructure. In this paper, we focus on Platform as a Service (PaaS) Software (SaaS) clouds, where users submit their workflow applications in order to be executed within strict timing constraints. It is assumed that target platform based multi-tenant approach, different may share same virtual machines. We propose list...
To address the challenges of imbalanced multi-class datasets typically used for rare event detection in critical cyber-physical systems, we propose an optimal, efficient, and adaptable mixed integer programming (MIP) ensemble weighting scheme. Our approach leverages diverse capabilities classifier on a granular per class basis, while optimizing weights classifier-class pairs using elastic net regularization improved robustness generalization. Additionally, it seamlessly optimally selects...
Summary Software as a Service (SaaS) cloud computing has emerged an attractive platform to tackle various problems of the traditional software distribution model, such requirement acquire and maintain expensive hardware infrastructure. SaaS, however, involves many challenges, mainly due heterogeneity multitenancy underlying host environment, well nature applications executed on platforms. Applications are usually bags‐of‐tasks, consisting independent component tasks that can be in any order,...
Extreme scale parallel computing systems will have tens of thousands optionally accelerator-equiped nodes with hundreds cores each, as well deep memory hierarchies and complex interconnect topologies. Such Exascale provide hardware parallelism at multiple levels be energy constrained. Their extreme the rapidly deteriorating reliablity their components means that exhibit low mean-time-between-failure values. Furthermore, existing programming models already require heroic optimisation efforts...
As cloud services become more ubiquitous, green computing attracts significant attention from both academia and industry. Towards this direction, in paper we propose an energy-aware heuristic for the scheduling of real-time workflow applications a environment. Our approach utilizes per-core Dynamic Voltage Frequency Scaling (DVFS) on underlying heterogeneous multi-core processors approximate computations, order to fill schedule gaps. goal is provide timeliness energy efficiency by trading...
One of the major challenges in a Software as Service (SaaS) cloud, is fault-tolerant and cost-effective scheduling execution end-user applications within strict time constraints, order to provide results acceptable quality. Towards this direction, we investigate performance strategies for fine-grained parallel SaaS cloud presence transient software failures, which may occur during applications. We combine compare by simulation different techniques that incorporate application-directed...
As the distributed resources required for processing of High Performance Computing (HPC) applications are becoming larger in scale and computational capacity, their energy consumption has become a major concern. Therefore, there is growing focus from both academia industry on minimization carbon footprint resources, especially through efficient scheduling workload. In this paper, technique proposed energy-aware bag-of-tasks with time constraints large-scale heterogeneous system. Its...
Distributed real-time systems play an increasingly vital role in our society. The most important aspect of such is the scheduling algorithm, which must guarantee that every job system will meet its deadline, providing high-quality (precise) results. In this paper we evaluate by simulation performance strategies for parallel jobs (gangs) a homogeneous distributed with possible software faults. For each policy provide alternative version allows imprecise computations. We propose metric...
In the highly competitive business environment of Software as a Service (SaaS) clouds, Quality (QoS) and fair pricing are paramount importance for differentiating between similar cloud providers. such platforms, workload computational demand variability may have significant impact on system performance thus provider's Level Agreement (SLA) commitments. Towards this direction, in paper we consider SaaS with multi-tier SLA that focuses billing end-users, according to provided level QoS. The...
Summary Cloud bursting is a concept originating from the hybrid cloud computing paradigm. During workload spikes, local resources of private are supplemented by in public cloud. This technique could also be applied fog environment, order to handle fluctuations, offloading applications Toward this direction, article, we propose strategy for utilization supplementary resources, assist processing Internet Things workflow jobs that arrive dynamically environment. As involves higher data transfer...
Summary As the adoption of Software as a Service (SaaS) cloud computing continues to gain momentum, arising challenges scheduling parallel applications on such platforms need be addressed. Due complexity and fine‐grained parallelism workload, well multi‐tenancy underlying host environment, end‐user are usually prone transient software failures. Therefore, fault tolerance is one most crucial aspects in SaaS clouds. It achieved through application‐directed checkpointing. However, selecting an...