PaperlayerView exact revision on arXiv

Source-linked AI summary

Survey on Aerial Radio Access Networks: Toward a Comprehensive 6G Access Infrastructure

Nhu-Ngoc Dao, Quoc-Viet Pham, Ngo Hoang Tu, Tran Thien Thanh, Vo Nguyen Quoc Bao, Demeke Shumeye Lakew, Sungrae Cho

arXiv:2102.07087v2cs.NIeess.SP

TL;DR

ARANs address limited support for mobile services in underserved areas by complementing terrestrial networks with flexible airborne access infrastructure. This survey synthesizes ARAN architectures, system models, enabling technologies, applications, and research challenges toward 6G access infrastructure.

  • Problem

    ARAN design must address battery-capacity limitations and incomplete modeling of transmission propagation, particularly beyond sub-6G GHz radio bands.

  • Method

    The paper surveys ARAN literature, derives a reference architecture from 5G and STIN standards, and reviews system models, enabling technologies, applications, and research challenges.

  • Results

    The survey identifies dynamic mobility, energy replenishment, operational management, and data delivery as central ARAN directions, with mobile UAV deployment reducing average total transmission power by 45%.

  • Takeaways & Limitations

    ARANs provide a framework for flexible 6G access infrastructure, while intelligent radio and federated learning are envisioned to coordinate adaptive networking behavior.

  • Takeaways & Limitations

    Jointly optimizing charging efficiency and communication objectives under user mobility and QoS requirements remains insufficiently investigated.

Abstract

from arXiv · show

Current network access infrastructures are characterized by heterogeneity, low latency, high throughput, and high computational capability, enabling massive concurrent connections and various services. Unfortunately, this design does not pay significant attention to mobile services in underserved areas. In this context, the use of aerial radio access networks (ARANs) is a promising strategy to complement existing terrestrial communication systems. Involving airborne components such as unmanned aerial vehicles, drones, and satellites, ARANs can quickly establish a flexible access infrastructure on demand. ARANs are expected to support the development of seamless mobile communication systems toward a comprehensive sixth-generation (6G) global access infrastructure. This paper provides an overview of recent studies regarding ARANs in the literature. First, we investigate related work to identify areas for further exploration in terms of recent knowledge advancements and analyses. Second, we define the scope and methodology of this study. Then, we describe ARAN architecture and its fundamental features for the development of 6G networks. In particular, we analyze the system model from several perspectives, including transmission propagation, energy consumption, communication latency, and network mobility. Furthermore, we introduce technologies that enable the success of ARAN implementations in terms of energy replenishment, operational management, and data delivery. Subsequently, we discuss application scenarios envisioned for these technologies. Finally, we highlight ongoing research efforts and trends toward 6G ARANs.

I. INTRODUCTION

The paper motivates aerial radio access networks (ARANs) as a complement to terrestrial infrastructure for extending mobile Internet access toward comprehensive 6G coverage. It defines ARANs as multitier, hierarchical aerial access systems and surveys their role, architecture, technologies, applications, and research challenges.

  • Motivation: 5G infrastructure emphasizes connectivity, throughput, latency, reliability, energy efficiency, and mobility, but does not comprehensively address serviceability in underserved or disrupted areas.Event-based communications can overload existing infrastructure, while disasters can damage or destroy network infrastructure even when communication remains critical.
  • Motivation: Aerial networking offers mobility, improved coverage, better channel access, and higher likelihood of line-of-sight propagation than terrestrial networks.UAVs can be deployed quickly and flexibly in three dimensions for multiple wireless-system purposes.
  • ARAN concept: ARANs provide radio access from the sky through aerial base stations, with aircraft, airships, UAVs, drones, balloons, and airplanes as typical platforms.Backhaul may use miniaturized satellites and terrestrial macro base stations; unifying these systems yields a multitier hierarchical reference model for 6G research.
  • Related work: The survey addresses ARANs within broader 6G visions that integrate terrestrial, airborne, and satellite communications into comprehensive access infrastructures.Prior work covers 6G visions, rural connectivity, unified three-dimensional coverage, AI-enabled 6G technologies, and heterogeneous or specialized RAN architectures.
  • Survey scope: Existing aerial-communication surveys are limited or topic-specific, motivating this comprehensive survey of ARAN architecture, system models, enabling technologies, applications, and research challenges.The surveyed topics include propagation, energy consumption, latency, mobility, energy refills, operational management, data delivery, and three application categories.

II. Network design

The paper organizes ARAN network design as a structured survey of architecture, system models, enabling technologies, applications, and research challenges. It presents ARANs as a reference framework spanning technical foundations and future development needs.

  • System model: The system-model analysis covers transmission propagation, energy consumption, communication latency, and system mobility.Propagation modeling addresses path loss in three-dimensional wireless environments, while energy analysis includes transmission, computation, motion, and hovering.
  • Enabling technologies: Enabling technologies are reviewed across energy replenishment, operational management, and data delivery.The scope includes charging approaches, network softwarization, mobile cloudization, data mining, frequency spectrum, communication protocols, and multiaccess schemes.
  • Application scenarios: Application scenarios are classified as event-based, scheduled, and permanent communications, with their requirements and common problems examined.The survey considers how ARAN advances can support these scenarios in 6G contexts.
  • Research challenges: Research challenges include intelligent radio, extremely high spectrum exploitation, network stability, security, privacy, and evaluation tools.The paper presents the survey as a reference framework that taxonomizes ARAN foundations and derives lessons from related-work analyses.
  • Network design: The survey positions ARANs within comprehensive 6G infrastructure and analyzes their system architecture and reference model.The network-design discussion provides a high-level account of ARAN positions, roles, and relationships with other infrastructure elements.

A. ARANs in 6G Access Infrastructure

ARANs unify low-altitude, high-altitude, and low-Earth-orbit communication systems into a layered aerial access infrastructure for diverse mobile and terrestrial users. Their architecture connects aerial access points to terrestrial core networks through front-end and back-end interfaces while adopting a 5G-based reference model.

  • ARANs in 6G Access Infrastructure: ARANs comprise LAP, HAP, and LEO systems serving high-altitude and terrestrial users across a large, dynamic coverage space.LAP communications occupy 0–10 km, HAP communications 20–50 km, and LEO communications 500–1500 km above sea level.
  • ARANs in 6G Access Infrastructure: ARANs unify related aerial networking classes whose partial tiers previously had weak and asynchronous interconnections.The review identifies DA-RANs, FlyRANs, and STINs as related classes.
  • System Architecture: The main segment is a cross-tier infrastructure shared across LAP, HAP, and LEO altitudes, complemented by front-end access and back-end core-network interfaces.Front-end components gather user connections, while the back end bridges ARAN infrastructure to terrestrial core networks.
  • System Architecture: LAP drones and UAVs provide direct aerial or terrestrial connectivity and use satellite or terrestrial links for backhaul transmission.LAP aerial base stations can connect to LEO systems or terrestrial macro base stations.
  • System Architecture: The ARAN reference model jointly adopts 3GPP and ETSI standards, treating ABSs as 5G gNB-functional components and LEO systems as trusted non-3GPP components.The model abstracts most 5G core components except the UPF, which contacts core networks and other systems.

D. Fundamental Features

ARANs distinguish themselves through layered aerial networking that supports broad coverage, adaptability, resilience, simultaneous delivery, and rapid topology formation. The survey analyzes these features alongside propagation, energy, latency, and mobility considerations.

  • Fundamental Features: ARANs provide ubiquitous service through LEO coverage, networking backup and resilience, and emergency broadcast capabilities.These advantages address situations where terrestrial infrastructure has limited capacity.
  • Fundamental Features: Dynamic LAP and HAP topologies and overlays across LAP, HAP, and LEO systems let ARANs adapt to end-user requirements on the ground and in the air.The feature is characterized as mobility.
  • Fundamental Features: Operating at multiple altitudes improves service availability across terrains and reduces exposure to disasters that can disrupt terrestrial infrastructure.The paper names mountains, seas, and deserts as example terrains.
  • Fundamental Features: Multitier systems can self-organize to deliver information-centric services through simultaneous multicast and broadcast streams across aerial and terrestrial locations.These streams use various wireless access technologies.
  • Fundamental Features: Hierarchical overlays and aerial ad hoc interlinks enable ARANs to establish scalable local topologies without service interruptions.The survey treats transmission propagation, energy consumption, latency, and mobility as four key system aspects.

1) Deterministic Models:

Deterministic propagation models represent aerial channels using environmental information and ray-tracing analysis, while related channel studies combine empirical, analytical, stochastic, and geometric approaches. Accurate channel characterization supports waveform, allocation, modulation, antenna, and interference design.

  • 1) Deterministic Models:: Deterministic channel models use propagation-environment information, including terrain topography and buildings or obstacles, and commonly employ ray-tracing software.Examples include air-to-ground path-loss and shadowing analysis in urban areas.
  • 1) Deterministic Models:: Free-space path loss is calculated from distance and carrier frequency through the Friis transmission equation.Distance is measured in 3D space between the ABS and ground user, while frequency is expressed in MHz.
  • 1) Deterministic Models:: Ground-to-air path-loss studies model total loss using terrestrial path loss and aerial excess path loss components in environment-specific settings.The cited suburban model makes aerial excess path loss dependent on channel conditions.
  • 1) Deterministic Models:: Channel-model studies report coherence bandwidths of 92 and 3846 kHz for corresponding delay resolutions, informing payload-symbol selection in frequency-selective LAP systems.The comparison concerns residential and mountainous desert scenarios.
  • 1) Deterministic Models:: Geometric-based stochastic models capture spatio-temporal characteristics and support analytical expressions for measures such as coverage radius and channel capacity.A cited 3-D MIMO model addresses LAP nonstationarity using time-varying arrival and departure angles.

1) Energy Consumption in Hover and Vertical Movement:

Aerial energy models account for vehicle, flight, payload, weather, and aerodynamic factors across hovering, vertical, and horizontal movement. Hovering models relate power to weight and rotor characteristics, while flight models incorporate lift-induced and parasitic drag.

  • 1) Energy Consumption in Hover and Vertical Movement:: Hovering-power models relate UAV consumption to total weight, including frame, battery, and payload weight.The rotor thrust is defined as T = (M + m)g, with M as frame weight and m as battery and payload weight.
  • 1) Energy Consumption in Hover and Vertical Movement:: Multirotor models assume each rotor bears an equal share of frame, battery, and payload weight, then aggregate rotor power into total copter power.The total-power expression uses N, the total number of rotors.
  • 1) Energy Consumption in Hover and Vertical Movement:: Linear approximations separate power for bearing battery and payload weight from power required for hovering, with takeoff and landing set equal to hovering power.This is an explicit simplification used in the cited models.
  • 1) Energy Consumption in Hover and Vertical Movement:: Drone flight power must overcome lift-induced and parasitic drag, whose models depend on air density, velocity, lift, wingspan, drag coefficient, and wing area.The paper distinguishes fixed-wing power models from rotary-wing models.

C. Latency Analysis

ARAN latency analysis covers end-to-end delay components, low-latency aerial services, and deployment or computation strategies that affect latency. Existing studies generally report lower latency with appropriate ABS deployment than with fixed baselines.

  • Latency Analysis: Aerial caching can reduce latency by serving users directly and avoiding requests to remote servers and network bottlenecks.
  • Latency Analysis: End-to-end latency is approximated as twice the sum of radio, backhaul, core, and transport latencies.Radio latency includes queuing, frame alignment, transmission, and processing delays.
  • Latency Analysis: Joint optimization of ABS trajectories and user associations achieved lower latency than baseline schemes with fixed deployment.
  • Latency Analysis: MEC studies report a trade-off between energy consumption and task completion time, with longer completion times consuming less energy.
  • Latency Analysis: Completion time is reduced when the number of UAVs increases, with UAVs serving jointly as computing servers and relays.

2) Trajectory Schedule:

ARAN trajectory scheduling uses mobility and deployment control to improve data collection, energy efficiency, and service performance. The literature also identifies practical trajectory constraints and future 6G modeling gaps.

  • Trajectory Schedule: Trajectory optimization can improve line-of-sight links, data collection efficiency, and the energy efficiency of mobile devices.
  • Trajectory Schedule: Unconstrained UAV trajectories were energy inefficient regardless of the design objective, motivating constraints on locations, velocity, and altitude.
  • Trajectory Schedule: Joint UAV scheduling and three-dimensional trajectory optimization can maximize the minimum data collection rate while scheduling one sensor at each time instance.
  • Summary and Discussion: ARAN analysis covers transmission propagation, energy consumption, latency, and mobility, while existing propagation studies mainly focus on sub-6G GHz bands.
  • Summary and Discussion: Latency analysis remains concentrated on typical 5G scenarios and requires study in future 6G settings such as massive URLLC and zero-touch services.
  • Summary and Discussion: Dynamic configuration and adaptive control can significantly improve system performance compared with conventional terrestrial or stationary RANs.

1) Charging Station:

ARAN charging strategies include conventional stations, tethered systems, and near-field wireless charging, while energy harvesting supports airborne endurance. These approaches involve trade-offs among coverage, mobility, charging range, efficiency, and communication performance.

  • Charging Station: Charging-station approaches comprise traditional charging stations, tethered technology, and near-field wireless charging.
  • Tethered Technology: Tethered UAVs can achieve 100% available time, while untethered UAVs show approximately 70% mission availability and 30% recharging time.
  • Tethered Technology: A 120-meter tether can extend coverage probability by up to 30% compared with untethered UAVs, but tether length and inclination constraints limit mobility.
  • Near-field Wireless Charging: Inductive charging operates in the kHz band over generally 20 cm and can reach up to 90% efficiency within 17.5–26.5 cm.
  • Near-field Wireless Charging: Magnetic-resonance charging operates in the MHz band and has been used in UAV charging stations to address endurance problems.
  • Energy Harvesting: Higher altitude can increase harvested energy, but it also increases communication path loss, creating an energy–communication trade-off.
  • Energy Harvesting: Solar energy harvesting requires jointly considering harvested energy, aerodynamic consumption, onboard storage, and ground-user QoS.

3) RF Wireless Charging:

RF wireless charging supports two complementary ARAN roles: replenishing airborne base stations and supplying power to low-power terrestrial devices. Its received power depends on transmitter power, wavelength, antenna gains, and distance, while reported charging can reach several tens of kilometers.

  • RF wireless charging: RF wireless charging operates from 300 MHz to 300 GHz and estimates received power using transmitter power, wavelength, antenna gains, and distance.The paper identifies this approach as far-field wireless charging and relates the estimate to the Friis formula.
  • RF wireless charging: Charging distances can reach several tens of kilometers, with power inversion efficiency reported up to 84% at 5.8 dBm cumulative received RF power.The paper defines energy efficiency as usable DC output power divided by received RF power.
  • Application scenarios: ARAN studies apply RF wireless charging with ABSs either as powered devices recharged by terrestrial macro base stations or as wireless power supplies for users.These two scenarios frame the reviewed applications of RF wireless charging.
  • ABSs as powered devices: A micro-UAV charging design used an integrated rectifier antenna and approximately 5 W of transmit power at 5.9 GHz to replenish its batteries.This is presented as an example of ABSs receiving power from terrestrial infrastructure.
  • ABSs as power supply: Dynamic ABSs can energize many low-power devices while jointly optimizing charging efficiency and communication throughput through energy allocation and harvest-transmit-store strategies.The reviewed work considers ABSs as aerial power sources for terrestrial wireless sensor and user devices.

B. Operational Management

ARAN operational management combines network softwarization, mobile cloudization, and data mining to coordinate heterogeneous resources, support computation across tiers, and inform adaptive network decisions. The reviewed literature applies these pillars across UAV, aerial, and satellite systems.

  • Operational management: Operational management rests on network softwarization, mobile cloudization, and data mining as three foundational pillars.The paper presents these pillars as the core organization of the operational management plane.
  • Network softwarization: Network softwarization harmonizes network and computational resources across ARAN tiers through programmable management and centralized orchestration.It provides operators with enhanced control, situational awareness, and flexibility while supporting coordination across network nodes.
  • Network softwarization: An SDN-UAV architecture separates data and control planes, allowing a controller to use global UAV-group information when selecting flight routes.The cited implementation uses network programmability to control UAV behavior.
  • Mobile cloudization: Mobile cloudization distributes computing resources across ARAN tiers and supports on-demand allocation, including offloading intensive tasks to LAP/HAP platforms with MEC.Reviewed applications target energy overhead and execution-delay reduction, while fog-cloud cooperation addresses latency and power consumption.
  • Data mining: Data mining extracts useful patterns from node data for prediction and action decisions, while ML supports massive connections in dynamic, heterogeneous ARAN environments.The paper associates these techniques with energy reduction, security, workload sharing, bandwidth improvement, self-organization, and mobility analysis.
  • Data mining: Reinforcement learning has been used in LEO communications to configure satellite links for high throughput, low bit error rate, bandwidth optimization, and reduced power consumption.The cited work applies learning-based decisions to links between LEO constellations and ground stations.

C. Data Delivery

ARAN data delivery relies on wireless technologies, high-frequency links, URLLC protocols, and multiaccess schemes to connect heterogeneous aerial and terrestrial tiers. The review covers spectrum, reliability, throughput, interference management, and simultaneous access.

  • Wireless technologies: 5G NR and Wi-Fi support links among LAP/HAP ABSs and between ABSs and terrestrial or aerial users across licensed and unlicensed sub-6 and THz bands.The paper also reports ultra-high-speed wireless backhaul using three-dimensional beamforming for ABSs.
  • Spectrum: THz links for LEO communications use RIS to compensate for high path loss, improving signal-to-noise ratio at high carrier frequencies.The reviewed studies examine THz bands from 0.1–10 THz for aerial and satellite links.
  • Communication protocols: URLLC is reviewed as an enabler for ARANs because it addresses broad coverage, low latency, and ultra-reliable communication requirements toward 6G.Short-packet UAV-IoT communication uses dynamic spectrum sharing and short blocklengths to reduce latency and avoid interference and congestion.
  • Multiaccess: Massive MIMO, NOMA, and RIS are identified as principal multiaccess technologies for enhancing ARAN performance.These approaches target spatial multiplexing, spectrum efficiency, capacity, latency, throughput, and interference management.
  • Massive MIMO: Massive MIMO research addresses simultaneous UAV access and LEO scheduling, including antenna deployment, user grouping, Doppler compensation, and time-delay compensation.The reviewed LEO study reports significantly improved data rate, while UAV work examines ergodic-rate maximization through antenna spacing.
  • NOMA: NOMA studies target multiconnection for many terrestrial users and report improvements in LEO ergodic capacity, outage probability, and mutual information.UAV and HAP studies additionally investigate path-following and beamforming for rate and bit-error performance.
  • Summary and discussion: The review summarizes energy replenishment, operational management, and data delivery as three technological pillars for emerging ARANs.It highlights energy replenishment as the foremost limiting aspect and links operational technologies to control, interoperability, energy efficiency, and latency reduction.

V. APPLICATION SCENARIOS

ARAN applications are organized by networking duration and requirements into event-based, scheduled, and permanent communications. The reviewed examples emphasize temporary emergency response, predefined-path monitoring, and services enabled by heterogeneous aerial tiers.

  • Application classification: ARAN applications span wireless coverage expansion, aerial surveillance, precision agriculture, and commercial delivery across hierarchical network tiers.The paper classifies these applications by networking requirements and service duration.
  • Event-based communications: Event-based communications temporarily provide or enhance connectivity for short-duration scenarios such as disasters and search-and-rescue operations.These scenarios address situations where communication infrastructure is damaged, congested, or urgently needed.
  • Event-based communications: LEO satellite image analysis using deep learning supports disaster detection, rescue missions, relief coordination, and damage assessment.The cited system analyzes images collected from LEO satellites for disaster-related tasks.
  • Event-based communications: Cooperative device, IoT, UAV, cellular, mobile ad-hoc, and satellite networks support post-disaster communication, recovery, failure detection, and failure localization.The passage describes these networks as interlinked for emergency communication services.
  • Event-based communications: UAV swarms can locate targets and establish continuous communication with terrestrial personnel, with mission completion times reduced as more UAVs are deployed.This is presented as a multiobjective optimization approach for search-and-rescue operations.
  • Scheduled communications: Scheduled communications use predefined flight paths to provide services for a specified duration, including aerial surveillance and smart agriculture.Surveillance applications cover security, resource exploration, wildfire and oil-spill detection, and environmental monitoring.
  • Scheduled communications: Drone surveillance supports real-time monitoring, extended operating time, multimode operation, and observation of remote or inaccessible areas.Thermal-camera UAV systems are described for detecting hidden activities in dense forests and supporting reconnaissance.

2) Smart Agriculture:

ARAN applications span event-based, scheduled, and permanent communications, with studies covering aerial surveillance, smart farming, smart cities, and underserved-area connectivity. Smart agriculture combines UAVs, sensing, IoT, big data, machine learning, and satellite data for monitoring and timely support.

  • Smart Agriculture: Smart agriculture integrates IoT, big data, machine learning, and UAVs into farming operations, including crop monitoring and disease detection.
  • Smart Agriculture: Satellite and drone/UAV data support larger-scale precision agriculture monitoring through remote sensing and high-spatial-resolution land classification.
  • Smart Agriculture: 93% accuracy was attained by a tested land-monitoring model using multitemporal, high-spatial-resolution Landsat 8 data.
  • Permanent Communications: Permanent-communication research addresses smart cities, healthcare, intelligent transportation, data collection, and remote or isolated areas.
  • Underserved Areas: Underserved-area studies report up to 38% efficiency improvement and up to 37.5% delay reduction with UAV-assisted intermediate links.
  • Application Categories: ARAN application studies classify networking requirements into event-based, scheduled, and permanent communications.

VI. RESEARCH CHALLENGES

The survey identifies open ARAN challenges spanning intelligent radio, spectrum exploitation, network stability, security and privacy, and realistic simulation. Its conclusion synthesizes architecture, system models, enabling technologies, applications, and future research directions for 6G access infrastructure.

  • A. Intelligent Radio: ARANs must incorporate flexible federated learning while accommodating heterogeneous AI-chip capabilities, service requirements, traffic patterns, and mobility.
  • B. Extremely High Spectrum Exploitation: THz and visible-light communications face environmental attenuation, requiring adaptive propagation and channel estimation, narrow-beam beamforming, and possible RIS relaying.
  • C. Network Stability: Highly mobile 3-D topologies and intermittent connectivity complicate ARAN organization, requiring mobility-aware routing, load balancing, and backup routes.
  • D. Security and Privacy Issues: Security and privacy remain critical because shared links, line-of-sight exposure, limited aerial resources, mobility, and massive communications challenge protection.
  • E. Simulation Tools: Simulation evaluations often omit mechanical, aerodynamic, and environmental constraints, so results may not accurately reflect specific real environments.
  • Conclusion: The survey covers ARAN network design, system models, enabling technologies, applications, and research challenges toward comprehensive 6G access infrastructure.
Loading 2102.07087v2…