Athina Bourdena

Athina Bourdena

Auto-generated profile, based on publicly available information

Athina Bourdena

Department of Business Administration and Tourism, Hellenic Mediterranean University, Heraklion, Crete, Greece

Research Focus & Metrics

Academic Career Length

2Years (2025-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
24 March 2025 in IEEE Access
Verified
Journal Article
Open access
Andreas AndreouAndreas Andreou
Andreas Andreou
Constandinos MavromoustakisConstandinos Mavromoustakis
Constandinos Mavromoustakis
Evangelos MarkakisEvangelos Markakis
Evangelos Markakis
ABAthina Bourdena
Athina Bourdena
George MastorakisGeorge Mastorakis
George Mastorakis
The rapid advancement of Artificial Intelligence (AI) is reshaping industries and driving global innovation. However, the increasing complexity of AI models demands substantial data and computational resources, leading to significant energy consumption and environmental impact. This article explores the integration of quantum computing and end-to-end automation strategies in cloud-edge architectures. It proposes a hybrid quantum-classical AI framework that enhances training efficiency and reduces data and processing intensity by minimizing energy consumption. The framework leverages automated model orchestration, adaptive resource allocation, and intelligent data processing at the edge to improve system efficiency. In addition, it addresses ethical considerations, including privacy, fairness, and trustworthiness, to ensure alignment with human values. This approach significantly improves AI performance while fostering a sustainable and ethical AI ecosystem.

27 August 2025 in IEEE Access
Verified
Journal Article
Open access
Andreas AndreouAndreas Andreou
Andreas Andreou
Constandinos MavromoustakisConstandinos Mavromoustakis
Constandinos Mavromoustakis
George MastorakisGeorge Mastorakis
George Mastorakis
ABAthina Bourdena
Athina Bourdena
Evangelos MarkakisEvangelos Markakis
Evangelos Markakis
Semantic communications prioritize transmitting meaningful information over raw data in communication systems. However, these systems face significant optimization challenges, particularly concerning resource efficiency and fidelity due to the costly and delicate nature of photonic resources. Quantum computing offers promising solutions to these challenges through its unique capabilities, such as superposition and entanglement, within high-dimensional Hilbert spaces. This survey reviews the integration of quantum computing into semantic communications, tracing developments from foundational concepts to current advancements. It explores how quantum embeddings and machine learning techniques enhance semantic representation and transmission, enabling the encoding and processing of only relevant information. This approach addresses issues of polysemy and contextual variations in large datasets more effectively than classical methods. Key topics include the utilization of quantum probability and cognition in semantic analysis, optimization of quantum protocols for faster information retrieval, and the role of variational quantum circuits in improving computation latency, communication bandwidth, data privacy, and transmission delays. It also examines the implications for future communication systems like next-generation networks. It highlights the shift towards intelligent computing-intensive architectures and the support for advanced applications such as extended reality and holographic communications. Potential pathways are identified to achieve more efficient, secure, and intelligent data transmission, aiming to provide researchers and practitioners with a thorough understanding of this emerging field, outlining open research challenges and future directions to inspire further exploration and innovation.

27 February 2025 in IEEE Access
Verified
Journal Article
Open access
Andreas AndreouAndreas Andreou
Andreas Andreou
Constandinos MavromoustakisConstandinos Mavromoustakis
Constandinos Mavromoustakis
Evangelos MarkakisEvangelos Markakis
Evangelos Markakis
ABAthina Bourdena
Athina Bourdena
George MastorakisGeorge Mastorakis
George Mastorakis
The article explores a novel framework that combines Ceiling-Mounted Systems (CMS) and Internet of Medical Things (IoMT) devices to address critical challenges in healthcare data management. By leveraging the capabilities of IoMT devices for real-time data collection, the proposed CMS’s robust sensing, storage, and processing features support efficient resource allocation in hospital environments. The proposed approach achieves approximately 35% reduction in latency, a 25% improvement in energy efficiency, and a 40% decrease in Age of Information (AoI) compared to traditional frameworks. The multi-objective optimization problem minimizes energy consumption and latency while ensuring fairness and timely data collection, which is particularly critical in patient monitoring and time-sensitive diagnostics scenarios. Deep Reinforcement Learning (DRL) methods solve the resource allocation challenge under realistic constraints. Specifically, Twin-Delayed Deep Deterministic Policy Gradients (TD3) and Soft Actor-Critic (SAC) algorithms are adopted to optimize task scheduling and system decisions in dynamic, resource-constrained settings. Simulation results demonstrate that SAC achieves approximately 20% faster convergence and 15% better adaptability in dynamic hospital environments compared to TD3, making it more suitable for real-time healthcare applications. These findings underscore the benefits of integrating IoMT devices with the proposed CMS infrastructures to meet healthcare requirements, such as robust security, high reliability, and real-time responsiveness.

16 September 2025 in IEEE Access
Verified
Journal Article
Open access
Andreas AndreouAndreas Andreou
Andreas Andreou
Constandinos MavromoustakisConstandinos Mavromoustakis
Constandinos Mavromoustakis
George MastorakisGeorge Mastorakis
George Mastorakis
ABAthina Bourdena
Athina Bourdena
Evangelos MarkakisEvangelos Markakis
Evangelos Markakis
In indoor smart spaces, ensuring reliable IoT connectivity is a significant challenge. Walls, furniture, and other obstacles often create coverage gaps that disrupt wireless links. This work presents a system-level integration and validation study of a hybrid networking framework that combines static, mobility-enabled, and reconfigurable components to sustain connectivity in complex indoor environments. First, a 3-D Voronoi-based deployment of static IoT nodes provides an optimized baseline coverage map. Next, a set of mobility-enabled platforms, ceiling-rail robots and floor Automated Guided Vehicles (AGVs), form an adaptive layer that can reposition as needed to restore links or extend coverage around obstacles. In parallel, Reconfigurable Intelligent Surfaces (RIS) are strategically placed on walls or ceilings to dynamically redirect wireless signals, creating virtual line-of-sight paths and mitigating shadowing. A multi-objective Genetic Algorithm (GA) coordinates these elements, optimizing coverage, connectivity, and latency simultaneously. Across hospital, office, and warehouse layouts, the integrated system achieves on average 95.6% volumetric coverage and 100% connectivity. Latency and energy-efficiency results are reported together with 95% confidence intervals to highlight statistical robustness. The study quantifies the gains of a unified controller versus sequential or disjoint orchestration and provides sensitivity analyses under RIS quantization and shadowing effects. The contribution is therefore positioned as a practical, unified framework that couples planning, mobility, and RIS operation under non-ideal hardware and channel conditions.