This talk does not have an abstract.

The adoption of Unmanned Aerial Vehicles (UAVs) as mobile Base Stations is a promising solution to boost the capacity in hotspot areas and to provide energy aware services in different smart environments.

The adoption of UAV-BSs involves the planning of their missions over time with the aim to provide specific services to ground users, which includes the scheduling of recharging actions of each UAV-BS at ground sites. This is of particular interest in two rising future technologies: Heterogeneous Cellular Networks using 5G and beyond techniques, and Internet of Things. The former leverages the utilization of mmWave technology on UAVs that recently gained attention due to high available bandwidth. The bandwidth budget can be provided to support demanding services, such as uplink live video streaming, or for edge computing purposes, such as those required seamless user interaction with real and virtual objects throughout extended reality multimedia services. As for the IoT, transmitting the huge amount of sensing data through UAVs may on a side alleviate the cellular network from a massive data collection and on the other side entail less power at the IoT end devices.

In the literature, Artificial Intelligence and Machine Learning based UAV flight planning algorithms aimed at improving energy efficiency as well as service quality related metrics are provided. Specifically, several algorithms leverage Q-learning and Reinforcement learning algorithms by introducing rewards related to key Quality of Service (QoS) and users’ Quality of Experience (QoE) metrics.

We will explore the potential for using UAVs BS in 5G/6G HetNet as well in cyber-physical systems. Emphasis will be placed on recent results using the Q-learning approaches, considering also several exciting emergent research directions paving the way towards novel applications such as eXtended Realty Multimedia Communications, Smart Health, Internet of Everything.

Over the last few years, the many uses of Unmanned Aerial Vehicles (UAVs) have captured the interest of both the scientific and the industrial communities. A typical scenario consists in the use of UAVs for surveillance or target-search missions over a wide geographical area. In this case, it is fundamental for the command center to accurately estimate and track the trajectories of the UAVs by exploiting their periodic state reports. In this work, we design an ad hoc tracking system that exploits the Long Range Wide Area Network (LoRaWAN) standard for communication and an extended version of the Constant Turn Rate and Acceleration (CTRA) motion model to predict drone movements in a three-dimensional environment. Simulation results on a publicly available dataset show that our system can reliably estimate the position and trajectory of a UAV, significantly outperforming baseline tracking approaches.

Today, Unmanned Aerial Vehicles (UAVs) are widely used in many different scenarios including search, monitoring, inspection, and surveillance. To be able to transmit the sensor data from the UAVs to the destination reliably within tangible response times to the relevant content is crucial, especially for tactical use cases. In this paper, we propose network coded torrents (NECTOR) to leverage multiple network interfaces for resilient hybrid satellite communications (SatCom) and terrestrial networking for UAVs. NECTOR is significantly different from the state of the art multipath protocols such as multipath TCP (MPTCP). Unlike MPTCP, NECTOR is a UDP-based user-space application that is deployable without any change in the operating system. Furthermore, NECTOR does not require any additional packet scheduler, rate-adaptation or forward error correction. We present the design and implementation of NECTOR, and evaluate its performance compared to MPTCP. Our experimental results show that NECTOR provides goodput (up to 70%) higher than MPTCP with less signaling overhead of nearly 5.49 times.

To extend connectivity and guarantee data rates, drones require a pertinent choice of hovering locations to cope with limitations such as flight time and coverage. To this end, we provide analytic bounds on the requirements of connectivity extension for vehicular networks served by fixed eMBB infrastructure, where both vehicular networks and fixed telecommunication infrastructure are modeled using stochastic and fractal geometry as a macro model for urban environment, providing a unique perspective into the smart city. Namely, we prove that assuming $n$ mobile nodes (distributed according to a hyperfractal distribution of dimension dF and an average of ρ gNBs (of dimension dr if ρ=nθwith θ>dr/4 then the average fraction of mobile nodes not covered by a gNB tends to zero like O(n−(dF−2)dr(2θ−dr2)). Interestingly, we then prove that the number of drones needed to connect the isolated mobile nodes is asymptotically equivalent to the number of isolated mobile nodes.

Recent years, under the cooperation of 5G techniques which enables high rate and low delay communication, cellular-connected unmanned aerial vehicles (UAVs) have attracted tremendous attentions due to its broad applications. In this paper, we focus on a mobile edge computing (MEC) enabled cellular-connected UAV system, where a UAV carried a fixed amount of computation tasks is deployed to fly from an initial location to a final location and served by ground base stations (GBSs) in the presence of a ground eavesdropper. During the flight, the UAV can offload part of the tasks to GBSs for remote execution. We aim at minimize the total energy consumption of UAV to complete the task, including computation energy, communication energy and flight-propulsion energy, by jointly optimizing UAV trajectory, computation task allocation, UAV-GBS association and transmit power allocation. To tackle the non-convexity of the original problem, we resort to the block coordinate decent and successive convex approximation techniques through decomposing it into three sub-problems, which are alternately solved until the algorithm converges. Numerical results show that the proposed design significantly saves UAV's energy consumption for complete the given tasks.

Group-oriented/multicast services in cellular networks are actually gaining momentum as the number of connected Internet of Things (IoT) devices is constantly growing and the need can rise of delivering the same content at the same time. Different sensors, video cameras, robots, and general-purpose IoT devices can be organised into multicast groups to receive the multicast service over a Point-to-Multipoint (PTM) transmission, thus improving the system spectral efficiency. When the group-oriented content is available in the radio access network (RAN) segment, the network wakes up devices by means of paging to inform them about the upcoming multicast transmission. However, most of the battery-powered devices are not immediately available for paging. In fact, devices can be reached at specific paging opportunities which vary from device to device. In a highly dense deployment, many devices will inevitably share the same paging opportunity forming a paging group. In this work, we analyze the fundamental trade-off between length of paging interval and size of paging subgroup in affecting the time to wait for the group content delivery. Furthermore, the impact of different paging group size distributions on the system performance is investigated.

Edge computing is an evolution of cloud computing, which brings application hosting from centralized data-centers down to the network edge, closer to mobile users and IoT devices. In this paper we focus on scenarios where edge computing is implemented by way of drones which behave as mobile edge servers and provide storage and processing services to offload data and maintain ubiquitous connectivity. In this scenario, Servers (i.e., drones) behave as sellers of network functions and Users (i.e. IoT devices) act as buyers, and interact with each other. To model these interactions, we hereby present a Stackelberg game which describes the system dynamics; we also show that an optimal equilibrium and convergence point for the game exists, which can be reached by way of a simple learning algorithm.

Unmanned aerial vehicles (UAVs) have been increasingly utilized in various civilian and military applications such as remote sensing, border patrolling, disaster monitoring, and communication coverage extension. However, there are still prone to several cyber attacks such as GPS spoofing attacks, distributed denial-of-service (DDoS) attacks, and man-in-the-middle attacks to obtain their collected information or to enforce the UAVs to perform their requested actions which may damage the UAVs or their surrounding environment or even endanger the safety of human in the operation field. In this paper, we propose a trust monitoring mechanism in which a centralized unit (e.g. the ground station) regularly observe the behavior of the UAVs in terms of their motion path, their consumed energy, as well as the number of their completed tasks and measure a relative trust score for the UAVs to detect any abnormal behaviors in a real-time manner. Our simulation results show that the trust model can detect malicious UAVs, which can be under various cyber-security attacks such as flooding attacks, man-in-the-middle attacks, GPS spoofing attack in real-time.

Despite the increasing popularity of commercial usage of UAVs or drone-delivered services, their dependence on the limited-capacity on-board batteries hinders their flight-time and mission continuity. As such, developing in-situ power transfer solutions for topping-up UAV batteries have the potential to extend their mission duration. In this paper, we study a scenario where UAVs are deployed as base stations (UAV-BS) providing wireless Hotspot services to the ground nodes, while harvesting wireless energy from flying energy sources. These energy sources are specialized UAVs (Charger or transmitter UAVs, tUAVs), equipped with wireless power transmitting devices such as RF antennae. tUAVs have the flexibility to adjust their flight path to maximize energy transfer. With the increasing number of UAV-BSs and environmental complexity, it is necessary to develop an intelligent trajectory selection procedure for tUAVs so as to optimize the energy transfer gain. In this paper, we model the trajectory optimization of tUAVs as a Markov Decision Process (MDP) problem and solve it using Q-Learning algorithm. Simulation results confirm that the Q-Learning based optimized trajectory of the tUAVs outperforms two benchmark strategies, namely random path planning and static hovering of the tUAVs.