This position paper presents A4FN, an Agentic Artificial Intelligence (AI) architecture for intent-driven automation in Flying Networks (FNs) using Unmanned Aerial Vehicles (UAVs) as access nodes. A4FN leverages Generative AI and Large Language Models (LLMs) to enable real-time, context-aware network control via a distributed agentic system. It comprises two components: the Perception Agent (PA), which semantically interprets multimodal input – including imagery, audio, and telemetry data – from UAV-mounted sensors to derive Service Level Specifications (SLSs); and the Decision-and-Action Agent (DAA), which reconfigures the network based on inferred intents. A4FN embodies key properties of Agentic AI, including autonomy, goal-driven reasoning, and continuous perception-action cycles. Designed for mission-critical, infrastructure-limited scenarios such as disaster response, it supports adaptive reconfiguration, dynamic resource management, and interoperability with emerging wireless technologies. The paper details the A4FN architecture, its core innovations, and open research challenges in multi-agent coordination and Agentic AI integration in next-generation FNs.

Unmanned Aerial Vehicles (UAVs) increasingly enhance the Quality of Service (QoS) in wireless networks due to their flexibility and cost-effectiveness. However, optimizing UAV placement in dynamic, obstacle-prone environments remains a significant research challenge due to their complexity. Reinforcement Learning (RL) offers adaptability and robustness in such environments, proving effective for UAV positioning optimization.This paper introduces RLpos-3, a novel framework that integrates standard RL techniques and simulation libraries with Network Simulator 3 (ns-3) to facilitate the development and evaluation of UAV positioning algorithms. RLpos-3 serves as a supplementary tool for researchers, enabling the implementation, analysis, and benchmarking of UAV positioning strategies across diverse environmental conditions while meeting user traffic demands. To validate its effectiveness, we present use cases demonstrating RLpos-3’s performance in optimizing UAV placement under realistic conditions, such as urban and obstacle-rich environments.

The 6G paradigm and the massive usage of in-terconnected wireless devices introduced the need for flexible wireless networks. A promising approach lies in employing Mobile Robotic Platforms (MRPs) to create communications cells on-demand. The challenge consists in positioning the MRPs to improve the wireless connectivity offered. This is exacerbated in millimeter wave (mmWave), Terahertz (THz), and visible light-based networks, which imply the establishment of short-range, Line of Sight (LoS) wireless links to take advantage of the ultra-high bandwidth channels available. This paper proposes a solution to enable the obstacle-aware, autonomous positioning of MRPs and provide LoS wireless connectivity to communications devices. It consists of 1) a Vision Module that uses video data gathered by the MRP to determine the location of obstacles, wireless devices and users, and 2) a Control Module, which autonomously positions the MRP based on the information provided by the Vision Module. The proposed solution was validated in simulation and through experimental testing, showing that it is able to position an MRP while ensuring LoS wireless links between a mobile communications cell and wireless devices or users.

Unmanned Aerial Vehicles (UAVs) are increasingly used as wireless communications nodes, serving as Wi-Fi Access Points and Cellular Base Stations. To enable energy-efficient access networks, we previously introduced the Sustainable multi-UAV Performance-aware Placement (SUPPLY) algorithm, which focuses on the energy-efficient placement of UAVs as Flying Access Points (FAPs) to serve Ground Users (GUs). However, SUPPLY did not address the backhaul link. This paper presents the Simple Gateway Positioning (SGWP) solution, which optimizes the position of a Gateway (GW) UAV to ensure backhaul connectivity in a two-tier network. We integrate SUPPLY for FAP positioning with SGWP for GW placement and evaluate their combined performance under various scenarios involving different GUs' Quality of Service (QoS) requirements and positions. Our results demonstrate that SUPPLY and SGWP can be used jointly in a two-tier network with minimal performance degradation.

Unmanned Aerial Vehicles (UAVs) have emerged as adequate platforms to carry communications nodes, including Wi-Fi Access Points and cellular Base Stations. This has led to the concept of flying networks composed of UAVs as a flexible and agile solution to provide on-demand wireless connectivity anytime, anywhere. However, state of the art works have been focused on optimizing the placement of the access network providing connectivity to ground users, overlooking the backhaul network design. In order to improve the overall Quality of Service (QoS) offered to ground users, the placement of Flying Gateways (FGWs) and the size of the queues configured in the UAVs need to be carefully defined to meet strict performance requirements. The main contribution of this article is a traffic-aware gateway placement and queue management (GPQM) algorithm for flying networks. GPQM takes advantage of knowing in advance the positions of the UAVs and their traffic demand to determine the FGW position and the queue size of the UAVs, in order to maximize the aggregate throughput and provide stochastic delay guarantees. GPQM is evaluated by means of ns-3 simulations, considering a realistic wireless channel model. The results demonstrate significant gains in the QoS offered when GPQM is used.

Unmanned Aerial Vehicles (UAVs) acting as Flying Access Points (FAPs) are being used to provide on-demand wireless connectivity in extreme scenarios. Despite ongoing research, the optimization of UAVs' positions according to dynamic users' traffic demands remains challenging. We propose the Traffic-aware UAV Placement Algorithm (TUPA), which positions a UAV acting as FAP according to the users' traffic demands, in order to maximize the network utility. Using a DRL approach enables the FAP to autonomously learn and adapt to dynamic conditions and requirements of networking scenarios. Moreover, the proposed DRL methodology allows TUPA to generalize knowledge acquired during training to unknown combinations of users' positions and traffic demands, with no additional training. TUPA is trained and evaluated using network simulator ns-3 and ns3-gym framework. The results demonstrate that TUPA increases the network utility, compared to baseline solutions, increasing the average network utility up to 4x in scenarios with heterogeneous traffic demands.

The increasing complexity of recent Wi-Fi amendments is making optimal Rate Adaptation (RA) a challenge. The use of classic algorithms or heuristic models to address RA is becoming unfeasible due to the large combination of configuration parameters along with the variability of the wireless channel. We propose a simple Deep Reinforcement Learning approach for the automatic RA in Wi-Fi networks, named Data-driven Algorithm for Rate Adaptation (DARA). DARA is standard-compliant. It dynamically adjusts the Wi-Fi Modulation and Coding Scheme (MCS) solely based on the observation of the Signal-to-Noise Ratio (SNR) of the received frames at the transmitter. Our simulation results show that DARA achieves higher throughput when compared with Minstrel High Throughput (HT) and matches the performance of the Ideal Wi-Fi RA algorithm.

The ability to operate virtually anywhere and carry payload makes Unmanned Aerial Vehicles (UAVs) perfect platforms to carry communications nodes, including Wi-Fi Access Points (APs) and cellular Base Stations (BSs). This is paving the way to the deployment of flying networks that enable communications to ground users on demand. Still, flying networks impose significant challenges in order to meet the Quality of Experience expectations. State of the art works addressed these challenges, but have been focused on routing and the placement of the UAVs as APs and BSs serving the ground users, overlooking the backhaul network design. The main contribution of this paper is a centralized traffic-aware Gateway UAV Placement (GWP) algorithm for flying networks with controlled topology. GWP takes advantage of the knowledge of the offered traffic and the future topologies of the flying network to enable backhaul communications paths with high enough capacity. The performance achieved using the GWP algorithm is evaluated using ns-3 simulations. The obtained results demonstrate significant gains regarding aggregate throughput and delay.

The ability to operate anywhere, anytime, as well as their capability to hover and carry cargo on board make Unmanned Aerial Vehicles (UAVs) suitable platforms to act as Flying Gateways (FGWs) to the Internet. The problem is the optimal placement of the FGWs within the flying network, such that the Quality of Service (QoS) offered is maximized. The literature has been focused on optimizing the placement of the Flying Access Points (FAPs), which establish high-capacity small cells to serve the users on the ground, overlooking the backhaul network design, including the FGW placement. The FGW placement problem is exacerbated in highly dynamic flying networks, where the dynamic traffic demand and the movements of the users may induce frequent changes in the placement of the FAPs.The main contribution of this paper is a fast gateway placement (F-GWP) algorithm for flying networks that determines the optimal position of a FGW. With F-GWP, backhaul communications paths with high enough capacity are established between the FAPs and the FGW, in order to accommodate the traffic demand of the users on the ground. Simulation and experimental results show F-GWP is two orders of magnitude faster than its state of the art counterpart, while ensuring the same flying network performance.

Besides the large amount of traffic that radio access and backhaul networks need to accommodate, the interest in low-latency communications is emerging. The goal is to reliably transmit high bitrates through the network under controlled delays, thus enabling human and machine-oriented communications. A critical aspect that must be addressed is the latency introduced by network queues. The literature has been focused on studying the queue size in wired networks, but wireless networks bring up additional challenges due to their dynamic characteristics. The problem is exacerbated in high dynamic networks, such as Flying Networks (FNs) composed of Unmanned Aerial Vehicles (UAVs), which have emerged to provide communications anywhere, anytime. The main contribution of this paper is a Proactive Queue Management (PQM) solution for FNs with controlled topology. PQM takes advantage of the knowledge of the future FN topologies and the offered traffic to define in advance the queue size of the communications nodes over time, in order to maximize the throughput with stochastic delay guarantees. The FN performance achieved using PQM is evaluated by means of ns-3 simulations, showing gains regarding aggregate throughput and average delay.

The use of Autonomous Underwater Vehicles (AUVs) is increasingly seen as a cost-effective way to carry out underwater missions. Due to their long endurance and set of sensors onboard, AUVs may collect large amounts of data, in the order of Gbytes, which need to be transferred to shore. State of the art wireless technologies suffer either from low bitrates or limited range. Since surfacing may be unpractical, especially for deep sea operations, long-range underwater data transfer is limited to the use of low bitrate acoustic communications, precluding the timely transmission of large amounts of data. The use of data mules combined with short-range, high bitrate RF or optical communications has been proposed as a solution to overcome the problem.In this paper we describe the implementation and validation of UDMSim, a simulation platform for underwater data muling oriented systems that combines an AUV simulator and the Network Simulator 3 (ns-3). The results presented in this paper show a good match between UDMSim, a theoretical model, and the experimental results obtained by using an underwater testbed when no localization errors exist. When these errors are present, the simulator is able to reproduce the navigation of AUVs that act as data mules, adjust the throughput, and simulate the signal and connection losses that the theoretical model can not predict, but that will occur in reality. UDMSim is made available to the community to support easy and faster evaluation of data muling oriented underwater communications solutions, and enable offline replication of real world experiments.

The growing demand for broadband communications anytime, anywhere has paved the way to the usage of Unmanned Aerial Vehicles (UAVs) for providing Internet access in areas without network infrastructure and enhancing the performance of existing networks. However, the usage of Flying Multi-hop Networks (FMNs) in such scenarios brings up significant challenges concerning network routing, in order to permanently provide the Quality of Service expected by the users. The problem is exacerbated in crowded events, where the FMN may be formed by many UAVs to address the traffic demand, causing inter-flow interference within the FMN. Typically, estimating inter-flow interference is not straightforward and requires the exchange of probe packets, thus increasing network overhead.The main contribution of this paper is an inter-flow interference-aware routing metric, named I2R, designed for centralized routing in FMNs with controllable topology. I2R does not require any control packets and enables the configuration of paths with minimal Euclidean distance formed by UAVs with the lowest number of neighbors in carrier-sense range, thus minimizing inter-flow interference in the FMN. Simulation results show the I2R superior performance, with significant gains in terms of throughput and end-to-end delay, when compared with state of the art routing metrics.

Unmanned Aerial Vehicles (UAVs) acting as aerial Wi-Fi Access Points or cellular Base Stations are being considered to deploy on-demand network capacity in order to serve traffic demand surges or replace Base Stations. The ability to estimate the Quality of Service (QoS) for a given network setup may help in solving UAV placement problems. This paper proposes a Machine Learning (ML) based QoS estimator, based on convolutional neural networks, which estimates the QoS for a given network by considering the UAV positions, the user positions and their offered traffic. The ML-based QoS estimator represents a novel paradigm for estimating the QoS in aerial wireless networks. It provides fast and accurate estimations with reduced computational complexity. We demonstrate the usefulness and applicability of the proposed QoS estimator using the ideal UAV placement algorithm. Simulation results show the QoS estimator has an average prediction error lower than 5%.

The growth of Internet of Things (IoT) technologies has triggered the development of low-cost solutions characterised by low energy consumption and low complexity. To interconnect these devices, some wireless communications technologies including IEEE 802.11 and IEEE 802.15.4 have been used due to their deployment and management simplicity and high scalability. However, in scenarios where the devices are physically distant or there is a massive number of devices in a reduced area, cellular technologies such as 3rd Generation Partnership Project (3GPP) Narrowband-Internet of Things (NB-IoT) are seen as the solution.This paper proposes a network planning theoretical model for NB-IoT, named NB-IoT Deterministic Link Adaptation Model (NB-DLAM), which can be used to estimate Quality of Service (QoS) metrics such as Packet Delivery Ratio (PDR), transmission time, and throughput. NB-DLAM estimations were compared with simulation results, which show the accuracy of the proposed model.

To properly validate wireless networking solutions we depend on experimentation. Simulation very often produces less accurate results due to the use of models that are simplifications of the real phenomena they try to model. Networking experimentation may offer limited repeatability and reproducibility. Being influenced by external random phenomena such as noise, interference, and multipath, real experiments are hardly repeatable. In addition, they are difficult to reproduce due to testbed operational constraints and availability. Without repeatability and reproducibility, the validation of the networking solution under evaluation is questionable. In this paper, we show how the Trace-based Simulation (TS) approach can be used to accurately repeat and reproduce real experiments and, consequently, introduce a paradigm shift when it comes to the evaluation of wireless networking solutions. We present an extensive evaluation of the TS approach using the Fed4FIRE+ w-iLab.2 testbed. The results show that it is possible to repeat and reproduce real experiments using Network Simulator 3 (ns-3) trace-based simulations with more accuracy than in pure simulation, with average accuracy gains above 50%.

Autonomous Underwater Vehicles (AUVs) are widely used as a cost-effective mean to carry out underwater missions. During long-term missions, AUVs may collect large amounts of data that usually needs to be sent to shore. An AUV may have to travel several kilometers before reaching an area of interest near the seafloor, thus surfacing is unpractical for most cases. Long-range underwater communications rely mostly on acoustic communications, which are characterized by very low bitrates, thus making the transfer of large amounts of data too slow. GROW is a novel solution for long-range, high bitrate underwater wireless communications between a survey unit (e.g., deep sea lander, AUV) and a central station at surface. GROW combines AUVs as data mules, short-range high bitrate wireless RF or optical communications, and long-range low bitrate acoustic communications for control.In this paper we present the Underwater Data Muling Protocol (UDMP), a communications protocol that enables the control and the scheduling of the Data Mule Units within the GROW framework. Experimental results obtained using an underwater testbed show that the use of UDMP and data mules can outperform acoustic communications, achieving equivalent throughput up to 150 times higher within the typical range of operation of the latter.

Despite recent advances, always-on broadband Internet connectivity is still not available in Temporary Crowded Events (TCEs). To solve this problem, this paper envisions a novel concept named Traffic-Aware Multi-Tier Flying Network (TMFN). A TMFN consists of a mobile and physically reconfigurable network of Flying Mesh Access Points (FMAPs) and Gateways, which is able to dynamically reconfigure its topology according to the users' traffic demands - characterized by the users' positions and offered traffic. To implement this concept, a novel traffic-aware Network Planning (NetPlan) algorithm is proposed, which dynamically determines the FMAPs' coordinates and Wi-Fi cell ranges according to the users' traffic demands, in order to improve the TMFN's aggregate throughput, without compromising the overall coverage. Simulation results obtained in scenarios typically observed in TCEs demonstrate improved Quality of Service metrics, specifically the mean throughput, thus validating the proposed NetPlan algorithm.

The advent of small and low-cost Unmanned Aerial Vehicles (UAVs) is paving the way to use swarms of UAVs to perform missions such as aerial video monitoring and infrastructure inspection. Within a swarm, UAVs communicate by means of a Flying Multi-hop Network (FMN), which due to its dynamics induces frequent changes of network topology and quality of the links. Recently, UAVs have also been used to provide Internet access and enhance the capacity of existing networks in Temporary Events. This brings up additional routing challenges not yet addressed, in order to provide always-on and high capacity paths able to meet the Quality of Service expected by the users.This paper presents RedeFINE, a centralized routing solution for FMNs that selects high-capacity paths between UAVs and avoids communications disruptions, by defining in advance the forwarding tables and the instants they shall be updated in the UAVs; this represents a major step forward with respect to traditional routing protocols. The performance evaluation of RedeFINE shows promising results, especially regarding Throughput and Packet Delivery Ratio, when compared with state of the art routing solutions.

Enterprise Wi-Fi networks have been increasingly considering energy efficiency. In this paper, we present the Wi-Green project wherein we are investigating new techniques and innovative solutions that will allow the minimization of the energy consumption in Wi-Fi networks. In Wi-Green we will consider an enterprise network, in which there is equipment from different vendors, with different ages and different consumption profiles.

IEEE 802.11-based wireless multimedia sensor networks (WMSN) are a cost-effective and flexible solution for video monitoring scenarios. Yet, they suffer from three major problems: bad performance, throughput unfairness, and energy inefficiency. Several approaches have been considered to tackle these problems but they are too restrictive or complex. In this paper we propose a scheduling approach using FM as a control channel to address the energy inefficiency problem. By taking advantage of the FM radio characteristics - higher coverage and lower energy consumption than Wi-Fi, our proposed approach uses FM as an always-on point-to-multipoint control channel used to turn off the nodes' Wi-Fi radio interfaces when they are not needed to transmit, receive, or relay data. Numerical and simulation analysis shows that our proposed scheduling mechanism significantly reduces energy consumption, while preserving performance and fairness characteristics.

Typical sensor networks are formed by low-end, battery operated devices, which rely on low-energy communication technologies, such as Bluetooth, Zigbee and ANT+, due to their energy efficiency. On the other hand, sensor networks increasingly need to be connected to the Internet, which implies adaptations of the TCP/IP stack to fit such wireless technologies. These adaptations bring additional complexity and imply new hardware, thus deployments are cumbersome and sub-optimal. Conversely, Wi-Fi is ubiquitous, can be seamlessly integrated with TCP/IP, and is energy-efficient with the right configurations; yet, its usage is still uncommon in e-health scenarios. For these reasons, we argue that a TCP/IP over Wi-Fi approach should be followed in e-health sensor networks. We propose a novel cross-layer, context-aware network configuration mechanism, which monitors the user and networking contexts and optimizes the configuration of the TCP/IP protocol stack accordingly. Our approach enables seamless integration between e-health wireless sensor networks and the TCP/IP backbone, while improving energy efficiency and reliability.

IEEE 802.11-based Stub Wireless Mesh Networks (WMNs) are a cost-effective and flexible solution to extend wired network infrastructures. Yet, they suffer from two major problems: inefficiency and unfairness. A number of approaches have been proposed to tackle these problems, but they are too restrictive, highly complex, or require time synchronization and modifications to the IEEE 802.11 MAC. PACE is a simple multi-hop scheduling mechanism for Stub WMNs overlaid on the IEEE 802.11 MAC that jointly addresses the inefficiency and unfairness problems. It limits transmissions to a single mesh node at each time and ensures that each node has the opportunity to transmit a packet in each network-wide transmission round. Simulation results demonstrate that PACE can achieve optimal network capacity utilization and greatly outperforms state of the art CSMA/CA-based solutions as far as goodput, delay, and fairness are concerned.

Abstract Ubiquitous Internet access is becoming a major requirement for end‐users due to the increasing number of services and applications supported over the Internet. Extending the coverage of current Wi‐Fi infrastructures installed in companies, universities and cities, has been considered a solution to help in fulfilling this requirement, namely when it comes to wireless and nomadic Internet access. This paper describes and analyses a new and simple solution, called Wi‐Fi network Infrastructure eXtension (WiFIX), aimed at extending current Wi‐Fi infrastructures. WiFIX is based on standard IEEE 802.1D bridges and a single‐message protocol that is able to self‐organize the network, and it only requires software changes in IEEE 802.11 access points (APs); no changes to IEEE 802.11 stations are needed. Overhead analysis and experimental results show both the higher efficiency of the solution compared to the IEEE 802.11s draft standard and its good performance as far as data throughput, delay and packet loss are concerned. Copyright © 2010 John Wiley & Sons, Ltd.

Wireless Mesh Networks (WMNs), which feature infrastructureless broadband network configurations, are attracting attention as an elemental technology when it comes to the extension of current WLAN infrastructures. State of the art solutions addressing WMNs usually assume the use of omnidirectional antennas. In this paper we evaluate the performance improvements obtained when using directional antennas. By using simulations, we analyze the gains in terms of throughput, delay and fairness, considering a grid network topology used to extend an infrastructure network. Simulation results show that by changing the type of antenna in use from omnidirectional to directional, the average throughput of a WMN can increase about 56% and the average network delay can be reduced by approximately 40%, without compromising fairness.

The changes in the communication paradigm envisioned for future networks, with peer-to-peer/symmetric attachments gaining momentum and two IP (Internet protocol) versions coexisting, will pose new challenges to mobile communication networks. Traditional IP auto-configuration mechanisms will not work properly, since they were designed mostly having in mind a client-server/asymmetric attachment model, they assume a single IP version paradigm, and they target the auto-configuration of devices only. The 1ST ambient networks project has introduced a new concept - the ambient network - that enables handling every communication entity, either a single device or an entire network, as an ambient network (AN). This paper describes a new efficient mechanism, named basic connectivity (BC) mechanism, for auto-configuring IP connectivity between attaching ANs. A proof-of-concept prototype, experimental results, and theoretical analysis show that BC suites the future networking paradigm and represents a solution more efficient than the current trial-and-error mechanism for auto-configuring IP connectivity.

In the Next Generation Networks (NGNs) users will carry multiple devices forming cooperative networks known as Personal Area Networks (PANs). Some existing technologies enable this type of networks, such as Bluetooth or IEEE 802.15.4, but a unified framework capable of self-organizing them dynamically in a full heterogeneous environment populated by these and other technologies still has to be defined. Also, these networks are envisioned to be connecting dynamically to the Internet, and may use two IP versions and their autoconfiguration mechanisms.In this paper we propose a new framework, the Autoconfiguration and Self-management of Personal Area Networks (ASPAN), which enables the automatic and dynamic deployment of PANs in the heterogeneous environments envisioned for NGNs and handles the automatic and dynamic connection of a PAN to the global Internet.

There is a growing trend towards convergence of telecommunication and data networks in order to support a richer set of services and applications. At the same time, increasing diversity and density of network access technologies has made the goal of providing connectivity anytime and anywhere a real possibility. Another important development is the emergence of small, low-complexity user owned networks, such as personal area networks and body area networks. Dynamic interworking, also known as network composition, between networks of different types and sizes is essential in the push towards convergence, as well as to realize truly seamless connectivity between heterogeneous access networks. Dynamic interworking requires signalling between different elements of the control planes of the different networks in order to coordinate the control functions and resources of the networks concerned. In this paper, we present the generic ambient network signalling protocol suite to address the diverse signalling requirements for dynamic interworking of networks.

Next generation communication networks will be characterized by the coexistence of multiple technologies and user devices in an integrated fashion. The increasing number of devices owned by a single user will lead to a new communication paradigm: users owning multiple devices that form cooperative networks, and networks of different users that communicate with each other, e.g., acquiring Internet access through each other. In this communication scenario no user intervention should be required and technology should seamlessly adapt to the user's context, preferences, and needs. In this paper we address one of those scenarios, interworking between personal area networks, using legacy technologies and the Ambient Network and network composition concepts, herein explained. We argue that new functionalities should be introduced to enable effortless use of legacy technologies in such dynamic and heterogeneous environments.