arXiv subjects

Josep Mangues-Bafalluy

Publications and source records attributed to Josep Mangues-Bafalluy.

At least 19 recordsLinked to original sources

arXiv2026

Toward Experimentation-as-a-Service in 5G/6G: The Plaza6G Prototype for AI-Assisted Trials

This paper presents Plaza6G, the first operational Experiment-as-a-Service (ExaS) platform unifying cloud resources with next-generation wireless infrastructure. Developed at CTTC in Barcelona, Plaza6G integrates GPU-accelerated compute clusters, multiple 5G cores, both open-source (e.g., Free5GC) and commercial (e.g., Cumucore), programmable RANs, and physical or emulated user equipment under unified orchestration. In Plaza6G, the experiment design requires minimal expertise as it is expressed in natural language via a web portal or a REST API. The web portal and REST API are enhanced with a Large Language Model (LLM)-based assistant, which employs retrieval-augmented generation (RAG) for up-to-date experiment knowledge and Low-Rank Adaptation (LoRA) for continuous domain fine-tuning. Over-the-air (OTA) trials leverage a four-chamber anechoic facility and a dual-site outdoor 5G network operating in sub-6~GHz and mmWave bands. Demonstrations include automated CI/CD integration with sub-ten-minute setup and interactive OTA testing under programmable propagation conditions. Machine-readable experiment descriptors ensure reproducibility, while future work targets policy-aware orchestration, safety validation, and federated testbed integration toward open, reproducible wireless experimentation.

cs.NI↗

arXiv2025

Resource Orchestration and Optimization in 6G Extreme-edge Scenario

6G networks envision a pervasive service infrastructure spanning from centralized cloud to distributed edge and highly dynamic extreme-edge domains. This vision introduces significant challenges in orchestrating services over heterogeneous, volatile, and often mobile resources beyond traditional operator control. To address these challenges, this demo presents a 6G-ready orchestration architecture focused on resource prediction and service resilience at the extreme-edge. The proposed solution integrates (i) an AI/ML-based Infrastructure Status Prediction Module, (ii) a Monitoring System capable of handling large-scale, diverse telemetry, and (iii) a Decision Engine and Actuator that ensures proactive

cs.NI↗

arXiv2025

Trustworthy Inter-Provider Agreements in 6G Using a Privacy-Enabled Hybrid Blockchain Framework

Inter-provider agreements are central to 6G networks, where administrative domains must securely and dynamically share services. To address the dual need for transparency and confidentiality, we propose a privacy-enabled hybrid blockchain setup using Hyperledger Besu, integrating both public and private transaction workflows. The system enables decentralized service registration, selection, and SLA breach reporting through role-based smart contracts and privacy groups. We design and deploy a proof-of-concept implementation, evaluating performance using end-to-end latency as a key metric within privacy groups. Results show that public interactions maintain stable latency, while private transactions incur additional overhead due to off-chain coordination. The block production rate governed by IBFT 2.0 had limited impact on private transaction latency, due to encryption and peer synchronization. Lessons learned highlight design considerations for smart contract structure, validator management, and scalability patterns suitable for dynamic inter-domain collaboration. Our findings offer practical insights for deploying trustworthy agreement systems in 6G networks using privacy-enabled hybrid blockchains.

cs.NI↗

arXiv2025

Performance Analysis, Lessons Learned and Practical Advice for a 6G Inter-Provider DApp on the Ethereum Blockchain

This paper presents a multi-contract blockchain framework for inter-provider agreements in 6G networks, emphasizing performance analysis under a realistic Proof-of-Stake (PoS) setting on Ethereum's Sepolia testnet. We begin by quantifying Ethereum Virtual Machine (EVM)-based gas usage for critical operations such as provider registration, service addition, and SLA penalty enforcement, observing that cold writes and deep data structures can each inflate gas consumption by up to 20\%. We then examine block-level dynamics when multiple transactions execute concurrently, revealing that moderate concurrency (e.g., 30--50 simultaneous transactions) can fill blocks to 80--90\% of their gas limit and nearly double finalization times from around 15~seconds to over 30~seconds. Finally, we synthesize these insights into a practical design guide, demonstrating that flattening nested mappings, consolidating storage writes, and selectively timing high-impact transactions can markedly reduce costs and latency spikes. Collectively, our findings underscore the importance of EVM-specific optimizations and transaction scheduling for large-scale decentralized applications in 6G telecom scenarios. The implementation is available online.

cs.NI↗

arXiv2025

Blockchain for Federated Learning in the Internet of Things: Trustworthy Adaptation, Standards, and the Road Ahead

As edge computing gains prominence in Internet of Things (IoTs), smart cities, and autonomous systems, the demand for real-time machine intelligence with low latency and model reliability continues to grow. Federated Learning (FL) addresses these needs by enabling distributed model training without centralizing user data, yet it remains reliant on centralized servers and lacks built-in mechanisms for transparency and trust. Blockchain and Distributed Ledger Technologies (DLTs) can fill this gap by introducing immutability, decentralized coordination, and verifiability into FL workflows. This article presents current standardization efforts from 3GPP, ETSI, ITU-T, IEEE, and O-RAN that steer the integration of FL and blockchain in IoT ecosystems. We then propose a blockchain-based FL framework that replaces the centralized aggregator, incorporates reputation monitoring of IoT devices, and minimizes overhead via selective on-chain storage of model updates. We validate our approach with IOTA Tangle, demonstrating stable throughput and block confirmations, even under increasing FL workloads. Finally, we discuss architectural considerations and future directions for embedding trustworthy and resource-efficient FL in emerging 6G networks and vertical IoT applications. Our results underscore the potential of DLT-enhanced FL to meet stringent trust and energy requirements of next-generation IoT deployments.

cs.NI↗

arXiv2025

An Empirical Smart Contracts Latency Analysis on Ethereum Blockchain for Trustworthy Inter-Provider Agreements

As 6G networks evolve, inter-provider agreements become crucial for dynamic resource sharing and network slicing across multiple domains, requiring on-demand capacity provisioning while enabling trustworthy interaction among diverse operators. To address these challenges, we propose a blockchain-based Decentralized Application (DApp) on Ethereum that introduces four smart contracts, organized into a Preliminary Agreement Phase and an Enforcement Phase, and measures their gas usage, thereby establishing an open marketplace where service providers can list, lease, and enforce resource sharing. We present an empirical evaluation of how gas price, block size, and transaction count affect transaction processing time on the live Sepolia Ethereum testnet in a realistic setting, focusing on these distinct smart-contract phases with varying computational complexities. We first examine transaction latency as the number of users (batch size) increases, observing median latencies from 12.5 s to 23.9 s in the Preliminary Agreement Phase and 10.9 s to 24.7 s in the Enforcement Phase. Building on these initial measurements, we perform a comprehensive Kruskal-Wallis test (p < 0.001) to compare latency distributions across quintiles of gas price, block size, and transaction count. The post-hoc analyses reveal that high-volume blocks overshadow fee variations when transaction logic is more complex (effect sizes up to 0.43), whereas gas price exerts a stronger influence when the computation is lighter (effect sizes up to 0.36). Overall, 86% of transactions finalize within 30 seconds, underscoring that while designing decentralized applications, there must be a balance between contract complexity and fee strategies. The implementation of this work is publicly accessible online.

cs.NI↗

arXiv2023

Decentralized Energy Marketplace via NFTs and AI-based Agents

The paper introduces an advanced Decentralized Energy Marketplace (DEM) integrating blockchain technology and artificial intelligence to manage energy exchanges among smart homes with energy storage systems. The proposed framework uses Non-Fungible Tokens (NFTs) to represent unique energy profiles in a transparent and secure trading environment. Leveraging Federated Deep Reinforcement Learning (FDRL), the system promotes collaborative and adaptive energy management strategies, maintaining user privacy. A notable innovation is the use of smart contracts, ensuring high efficiency and integrity in energy transactions. Extensive evaluations demonstrate the system's scalability and the effectiveness of the FDRL method in optimizing energy distribution. This research significantly contributes to developing sophisticated decentralized smart grid infrastructures. Our approach broadens potential blockchain and AI applications in sustainable energy systems and addresses incentive alignment and transparency challenges in traditional energy trading mechanisms. The implementation of this paper is publicly accessible at \url{https://github.com/RasoulNik/DEM}.

cs.NI↗

arXiv2023

X-GRL: An Empirical Assessment of Explainable GNN-DRL in B5G/6G Networks

The rapid development of artificial intelligence (AI) techniques has triggered a revolution in beyond fifth-generation (B5G) and upcoming sixth-generation (6G) mobile networks. Despite these advances, efficient resource allocation in dynamic and complex networks remains a major challenge. This paper presents an experimental implementation of deep reinforcement learning (DRL) enhanced with graph neural networks (GNNs) on a real 5G testbed. The method addresses the explainability of GNNs by evaluating the importance of each edge in determining the model's output. The custom sampling functions feed the data into the proposed GNN-driven Monte Carlo policy gradient (REINFORCE) agent to optimize the gNodeB (gNB) radio resources according to the specific traffic demands. The demo demonstrates real-time visualization of network parameters and superior performance compared to benchmarks.

cs.NI↗

arXiv2023

A Multi-Agent Deep Reinforcement Learning Approach for RAN Resource Allocation in O-RAN

Artificial intelligence (AI) and Machine Learning (ML) are considered as key enablers for realizing the full potential of fifth-generation (5G) and beyond mobile networks, particularly in the context of resource management and orchestration. In this demonstration, we consider a fully-fledged 5G mobile network and develop a multi-agent deep reinforcement learning (DRL) framework for RAN resource allocation. By leveraging local monitoring information generated by a shared gNodeB instance (gNB), each DRL agent aims to optimally allocate radio resources concerning service-specific traffic demands belonging to heterogeneous running services. We perform experiments on the deployed testbed in real-time, showing that DRL-based agents can allocate radio resources fairly while improving the overall efficiency of resource utilization and minimizing the risk of over provisioning.

cs.NI↗

arXiv2023

Explanation-Guided Deep Reinforcement Learning for Trustworthy 6G RAN Slicing

The complexity of emerging sixth-generation (6G) wireless networks has sparked an upsurge in adopting artificial intelligence (AI) to underpin the challenges in network management and resource allocation under strict service level agreements (SLAs). It inaugurates the era of massive network slicing as a distributive technology where tenancy would be extended to the final consumer through pervading the digitalization of vertical immersive use-cases. Despite the promising performance of deep reinforcement learning (DRL) in network slicing, lack of transparency, interpretability, and opaque model concerns impedes users from trusting the DRL agent decisions or predictions. This problem becomes even more pronounced when there is a need to provision highly reliable and secure services. Leveraging eXplainable AI (XAI) in conjunction with an explanation-guided approach, we propose an eXplainable reinforcement learning (XRL) scheme to surmount the opaqueness of black-box DRL. The core concept behind the proposed method is the intrinsic interpretability of the reward hypothesis aiming to encourage DRL agents to learn the best actions for specific network slice states while coping with conflict-prone and complex relations of state-action pairs. To validate the proposed framework, we target a resource allocation optimization problem where multi-agent XRL strives to allocate optimal available radio resources to meet the SLA requirements of slices. Finally, we present numerical results to showcase the superiority of the adopted XRL approach over the DRL baseline. As far as we know, this is the first work that studies the feasibility of an explanation-guided DRL approach in the context of 6G networks.

cs.NI↗

arXiv2022

Video on Demand Streaming Using RL-based Edge Caching in 5G Networks

Edge caching can significantly improve the 5G networks' performance both in terms of delay and backhaul traffic. We use a reinforcement learning-based (RL-based) caching technique that can adapt to time-location-dependent popularity patterns for on-demand video contents. In a private 5G, we implement the proposed caching scheme as two virtual network functions (VNFs), edge and remote servers, and measure the cache hit ratio as a KPI. Combined with the HLS protocol, the proposed video-on-demand (VoD) streaming is a reliable and scalable service that can adapt to content popularity.

cs.NI↗

arXiv2022

Vertical-oriented 5G platform-as-a-service: user-generated content case study

5G realizes an impactful convergence, where Network Functions Virtualization (NFV) and cloud-native models become fundamental for profiting from the unprecedented capacity offered at the 5G Radio Access Network (RAN). For providing scalability and automation management over resources in 5G infrastructure, cloud-native and Platform as a service (PaaS) are proposed as solutions for paving the way for vertical applications in 5G. This paper leverages cloud-native models, PaaS, and virtual testbed instances to provide key platform provisioning and service life-cycle management features to a selected User Generated Content (UGC) scenario in multimedia applications. Specifically, this article and results show how service-level telemetry from a UGC cloud-native application is used to automatically scale system resources across the NFV infrastructure.

cs.NI↗

arXiv2022

Intent-Based Orchestration for Application Relocation in a 5G Cloud-native Platform

The need of mobile network operators for cost-effectiveness is driving 5G and beyond networks towards highly flexible and agile deployments to adapt to dynamic and resource-constrained scenarios while meeting a myriad of user network stakeholders' requirements. In this setting, we consider that zero-touch orchestration schemes based on cloud-native deployments equipped with end-to-end monitoring capabilities provide the necessary technology mix to be a solution candidate. This demonstration, built on top of an end-to-end cloud-native 5G experimental platform with over-the-air transmissions, shows how dynamic orchestration can relocate container-based end-user applications to fulfil intent-based requirements. Accordingly, we provide an experimental validation to showcase how the platform enables the desired flexible and agile 5G deployments.

cs.NI↗

arXiv2022

Cloud-native 5G experimental platform with over-the-air transmissions and end-to-end monitoring

5G represents a revolutionary shift with respect to previous generations given its design centered on network softwarization. Within such a change of paradigm, cloud-native solutions are widely regarded as the future of vertical application development because of their enhanced flexibility and adaptability to complex and dynamic scenarios. In this context, we present an experimental framework with over-the-air transmissions that tackles two critical aspects for enhancing the lifecycle management of 5G and beyond networks: cloud-native deployments of 5G core network functions (NFs) and end-to-end monitoring. First, we deploy Open5GS and Prometheus-based monitoring as containerized network functions (CNFs) in a Kubernetes cluster spanning a multi-tier network with a multi-access edge computing (MEC) host. We then demonstrate the end-to-end monitoring system by showcasing via Grafana dashboards both infrastructure resources and radio metrics of two scenarios; one devoted to user plane function (UPF) re-selection and the other to user mobility.

cs.NI↗

arXiv2021

Multi-Provider NFV Network Service Delegation via Average Reward Reinforcement Learning

In multi-provider 5G/6G networks, service delegation enables administrative domains to federate in provisioning NFV network services. Admission control is fundamental in selecting the appropriate deployment domain to maximize average profit without prior knowledge of service requests' statistical distributions. This paper analyzes a general federation contract model for service delegation in various ways. First, under the assumption of known system dynamics, we obtain the theoretically optimal performance bound by formulating the admission control problem as an infinite-horizon Markov decision process (MDP) and solving it through dynamic programming. Second, we apply reinforcement learning to practically tackle the problem when the arrival and departure rates are not known. As Q-learning maximizes the discounted rewards, we prove it is not an efficient solution due to its sensitivity to the discount factor. Then, we propose the average reward reinforcement learning approach (R-Learning) to find the policy that directly maximizes the average profit. Finally, we evaluate different solutions through extensive simulations and experimentally using the 5Growth platform. Results confirm that the proposed R-Learning solution always outperforms Q-Learning and the greedy policies. Furthermore, while there is at most 9% optimality gap in the ideal simulation environment, it competes with the MDP solution in the experimental assessment.

cs.NI↗

arXiv2021

Near Optimal VNF Placement in Edge-Enabled 6G Networks

Softwarization and virtualization are key concepts for emerging industries that require ultra-low latency. This is only possible if computing resources, traditionally centralized at the core of communication networks, are moved closer to the user, to the network edge. However, the realization of Edge Computing (EC) in the sixth generation (6G) of mobile networks requires efficient resource allocation mechanisms for the placement of the Virtual Network Functions (VNFs). Machine learning (ML) methods, and more specifically, Reinforcement Learning (RL), are a promising approach to solve this problem. The main contributions of this work are twofold: first, we obtain the theoretical performance bound for VNF placement in EC-enabled6G networks by formulating the problem mathematically as a finite Markov Decision Process (MDP) and solving it using a dynamic programming method called Policy Iteration (PI). Second, we develop a practical solution to the problem using RL, where the problem is treated with Q-Learning that considers both computational and communication resources when placing VNFs in the network. The simulation results under different settings of the system parameters show that the performance of the Q-Learning approach is close to the optimal PI algorithm (without having its restrictive assumptions on service statistics). This is particularly interesting when the EC resources are scarce and efficient management of these resources is required.

cs.NI↗

arXiv2021

Model-Based Reinforcement Learning Framework of Online Network Resource Allocation

Online Network Resource Allocation (ONRA) for service provisioning is a fundamental problem in communication networks. As a sequential decision-making under uncertainty problem, it is promising to approach ONRA via Reinforcement Learning (RL). But, RL solutions suffer from the sample complexity issue; i.e., a large number of interactions with the environment needed to find an efficient policy. This is a barrier to utilize RL for ONRA as on one hand, it is not practical to train the RL agent offline due to lack of information about future requests, and on the other hand, online training in the real network leads to significant performance loss because of the sub-optimal policy during the prolonged learning time. This performance degradation is even higher in non-stationary ONRA where the agent should continually adapt the policy with the changes in service requests. To deal with this issue, we develop a general resource allocation framework, named RADAR, using model-based RL for a class of ONRA problems with the known immediate reward of each action. RADAR improves sample efficiency via exploring the state space in the background and exploiting the policy in the decision-time using synthetic samples by the model of the environment, which is trained by real interactions. Applying RADAR on the multi-domain service federation problem, to maximize profit via selecting proper domains for service requests deployment, shows its continual learning capability and up to 44% performance improvement w.r.t. the standard model-free RL solution.

cs.NI↗

arXiv2021

R-Learning Based Admission Control for Service Federation in Multi-domain 5G Networks

Service federation in 5G/B5G networks enables service providers to orchestrate network services across multiple domains where admission control is a key issue. For each demand, without knowing the future ones, the admission controller either determines the domain to deploy the demand or rejects it in order to maximize the long-term average profit. In this paper, at first, under the assumption of knowing the arrival and departure rates of demands, we obtain the optimal admission control policy by formulating the problem as a Markov decision process that is solved by the policy iteration method. As a practical solution, where the rates are not known, we apply the Q-Learning and R-Learning algorithms to approximate the optimal policy. The extensive simulation results show the learning approaches outperform the greedy policy, and while the performance of Q-Learning depends on the discount factor, the optimality gap of the R-Learning algorithm is at most 3-5% independent of the system configuration.

cs.NI↗