arXiv subjects

Panagiotis Trakadas

Publications and source records attributed to Panagiotis Trakadas.

8 recordsLinked to original sources

arXiv2026

DECOFFEE: Decentralized Reinforcement Learning for Time-critical Workload Offloading and Energy Efficiency across the Computing Continuum

The rapid proliferation of latency-sensitive and battery-constrained Internet-of-Things (IoT) applications has intensified the need for intelligent workload placement mechanisms across the Edge-Cloud computing continuum. In such environments, far-edge nodes must dynamically decide whether to execute workloads locally or offload them to neighboring nodes or the cloud, while accounting for execution delay, energy consumption, and strict timeout constraints. However, workload placement in large-scale distributed infrastructures is a highly dynamic and non-convex optimization problem due to stochastic arrivals, heterogeneous computing capacities, and time-varying network conditions. This paper proposes DECOFFEE, a decentralized reinforcement learning framework for time-critical workload offloading and energy-efficient operation across the computing continuum. The proposed multi-agent learning scheme jointly optimizes system delay, energy consumption, and workload drop rate through adaptive placement decisions. Each edge agent operates as an autonomous learning entity that derives an optimal policy from local system observations and predicted network conditions. The workload placement process is formulated as parallel Markov Decision Processes and solved using a Double Dueling Deep Q-Network (DQN) architecture enhanced with Long Short-Term Memory (LSTM) forecasting to anticipate future load conditions. Extensive simulations demonstrate that DECOFFEE and its variants consistently outperform conventional rule-based and heuristic placement strategies, achieving significant reductions in delay, energy consumption, and workload drop rate under varying traffic and network conditions.

cs.NI↗

arXiv2026

CARGO: Carbon-Aware Gossip Orchestration in Smart Shipping

Smart shipping operations increasingly depend on collaborative AI, yet the underlying data are generated across vessels with uneven connectivity, limited backhaul, and clear commercial sensitivity. In such settings, server-coordinated FL remains a weak systems assumption, depending on a reachable aggregation point and repeated wide-area synchronization, both of which are difficult to guarantee in maritime networks. A serverless gossip approach therefore represents a more natural approach, but existing methods still treat communication mainly as an optimization bottleneck, rather than as a resource that must be managed jointly with carbon cost, reliability, and long-term participation balance. In this context, this paper presents CARGO, a carbon-aware gossip orchestration framework for smart-shipping. CARGO separates learning into a control and a data plane. The data plane performs local optimization with compressed gossip exchange, while the control plane decides, at each round, which vessels should participate, which communication edges should be activated, how aggressively updates should be compressed, and when recovery actions should be triggered. We evaluate CARGO under a predictive-maintenance scenario using operational bulk-carrier engine data and a trace-driven maritime communication protocol that captures client dropout, partial participation, packet loss, and multiple connectivity regimes, derived from mobility-aware vessel interactions. Across the tested stress settings, CARGO consistently remains in the high-accuracy regime while reducing carbon footprint and communication overheads, compared to accuracy-competitive decentralized baselines. Overall, the conducted performance evaluation demonstrates that CARGO is a feasible and practical solution for reliable and resource-conscious maritime AI deployment.

cs.AI↗

arXiv2026

CollabORAN: A Collaborative rApp-xApp-dApp Control Architecture for Fairness-Adaptive Resource Sharing in O-RAN

The evolution of Open Radio Access Networks (O-RAN) enables programmable and intelligent control of radio resources through disaggregated architectures and open interfaces. However, existing solutions typically rely on isolated control loops and fail to jointly address end-to-end optimization objectives across multiple timescales. Thus, it remains a key challenge to functionally split optimization algorithms across timescale-specific O-RAN layers while complying with control loop latency specifications. This article proposes CollabORAN, a collaborative rApp-xApp-dApp hierarchical framework for dynamic and equitable spectrum sharing in O-RAN systems. CollabORAN leverages a nested control structure in which the rApp performs traffic-aware policy generation, the xApp executes interference-aware spectrum allocation via hypergraph-based PRB coloring, and the DU-level dApp enforces temporal fairness through fast scheduling. The proposed end-to-end closed-loop design enables coordinated optimization across minutes, seconds, and millisecond time scales. Simulation results demonstrate that CollabORAN significantly improves service fairness and reduces user starvation while maintaining efficient spectrum reuse in dense and dynamic network environments.

cs.NI↗

arXiv2026

A Dynamic Framework for Grid Adaptation in Kolmogorov-Arnold Networks

Kolmogorov-Arnold Networks (KANs) have recently demonstrated promising potential in scientific machine learning, partly due to their capacity for grid adaptation during training. However, existing adaptation strategies rely solely on input data density, failing to account for the geometric complexity of the target function or metrics calculated during network training. In this work, we propose a generalized framework that treats knot allocation as a density estimation task governed by Importance Density Functions (IDFs), allowing training dynamics to determine grid resolution. We introduce a curvature-based adaptation strategy and evaluate it across synthetic function fitting, regression on a subset of the Feynman dataset and different instances of the Helmholtz PDE, demonstrating that it significantly outperforms the standard input-based baseline. Specifically, our method yields average relative error reductions of 25.3% on synthetic functions, 9.4% on the Feynman dataset, and 23.3% on the PDE benchmark. Statistical significance is confirmed via Wilcoxon signed-rank tests, establishing curvature-based adaptation as a robust and computationally efficient alternative for KAN training.

cs.LG↗

arXiv2026

Interoperable rApp/xApp Control over O-RAN for Mobility-aware Dynamic Spectrum Allocation

Open Radio Access Networks (O-RAN) enable the disaggregation of radio access functions and the deployment of control applications across different timescales. However, designing interoperable control schemes that jointly exploit long-term traffic awareness and near-real-time radio resource optimization remains a challenging problem, particularly under dense multi-cell interference and heterogeneous service demands. This paper proposes an interoperable rApp/xApp-driven dynamic spectrum allocation (DSA) framework for O-RAN, based on a graph-theoretic formulation of physical resource block (PRB) assignment. The proposed architecture leverages a non-real-time radio intelligent controller (Non-RT RIC) rApp to predict aggregated traffic evolution and generate high-level spectrum policies at the minutes timescale, while a near-real-time RIC (Near-RT RIC) xApp constructs a user-centric conflict graph and performs fairness-aware PRB allocation at sub-second timescales. To mitigate persistent user starvation, a conflict-aware modified proportional fair (MPF) scheduling mechanism is applied, enabling controlled interference-free PRB time-sharing. Extensive simulation results demonstrate that the proposed framework significantly improves the PRB assignment success rate (above 90%) and service-share fairness (above 85%) across different channel configurations and user demands, while maintaining architectural separation and rApp/xApp interoperability in accordance with O-RAN principles.

cs.NI↗

arXiv2025

COMIX: Generalized Conflict Management in O-RAN xApps -- Architecture, Workflow, and a Power Control case

Open Radio Access Network (O-RAN) is transforming the telecommunications landscape by enabling flexible, intelligent, and multi-vendor networks. Central to its architecture are xApps hosted on the Near-Real-Time RAN Intelligent Controller (Near-RT RIC), which optimize network functions in real time. However, the concurrent operation of multiple xApps with conflicting objectives can lead to suboptimal performance. This paper introduces a generalized Conflict Management scheme for Multi-Channel Power Control in O-RAN xApps (COMIX), designed to detect and resolve conflicts between xApps. To demonstrate COMIX, we focus on two Deep Reinforcement Learning (DRL)-based xApps for power control: one maximizes the data rare across UEs, and the other optimizes system-level energy efficiency. COMIX employs a standardized Conflict Mitigation Framework (CMF) for conflict detection and resolution and leverages the Network Digital Twin (NDT) to evaluate the impact of conflicting actions before applying them to the live network. We validate the framework using a realistic multi-channel power control scenario under various conflict resolution policies, demonstrating its effectiveness in balancing antagonistic objectives. Our results highlight significant network energy savings achieved through the conflict management scheme compared to baseline CMF-free methods.

cs.NI↗

arXiv2022

A Survey on UAV-Aided Maritime Communications: Deployment Considerations, Applications, and Future Challenges

Maritime activities represent a major domain of economic growth with several emerging maritime Internet of Things use cases, such as smart ports, autonomous navigation, and ocean monitoring systems. The major enabler for this exciting ecosystem is the provision of broadband, low-delay, and reliable wireless coverage to the ever-increasing number of vessels, buoys, platforms, sensors, and actuators. Towards this end, the integration of unmanned aerial vehicles (UAVs) in maritime communications introduces an aerial dimension to wireless connectivity going above and beyond current deployments, which are mainly relying on shore-based base stations with limited coverage and satellite links with high latency. Considering the potential of UAV-aided wireless communications, this survey presents the state-of-the-art in UAV-aided maritime communications, which, in general, are based on both conventional optimization and machine-learning-aided approaches. More specifically, relevant UAV-based network architectures are discussed together with the role of their building blocks. Then, physical-layer, resource management, and cloud/edge computing and caching UAV-aided solutions in maritime environments are discussed and grouped based on their performance targets. Moreover, as UAVs are characterized by flexible deployment with high re-positioning capabilities, studies on UAV trajectory optimization for maritime applications are thoroughly discussed. In addition, aiming at shedding light on the current status of real-world deployments, experimental studies on UAV-aided maritime communications are presented and implementation details are given. Finally, several important open issues in the area of UAV-aided maritime communications are given, related to the integration of sixth generation (6G) advancements.

cs.NI↗

arXiv2019

Secure Open Federation of IoT Platforms Through Interledger Technologies -- The SOFIE Approach

The lack of interoperability among IoT platforms has led to a fragmented environment, where the users and society as a whole suffer from lock-ins, lack of privacy, and reduced functionality. This paper presents SOFIE, a solution for federating the existing IoT platforms in an open and secure manner using Distributed Ledger Technologies (DLTs) and without requiring modifications to the IoT platforms, and describes how SOFIE is used to enable two complex real life pilots: food supply chain tracking from field to fork and electricity distribution grid balancing with guided electrical vehicle (EV) charging. SOFIE's main contribution is to provide interoperability between IoT systems while also enabling new functionality and business models.

cs.CR↗