PaperlayerView exact revision on arXiv

Source-linked AI summary

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

Nikolaos Nomikos, Panagiotis K. Gkonis, Petros S. Bithas, Panagiotis Trakadas

arXiv:2209.09605v1cs.NIeess.SP

TL;DR

Maritime networks need ubiquitous connectivity, making efficient UAV integration an important research challenge. This survey organizes UAV-aided maritime communication techniques and architectures, and concludes that UAVs support maritime IoT and broadband services.

  • Problem

    How to efficiently integrate UAVs into maritime communication networks to achieve ubiquitous connectivity is a central challenge.

    Source: §I · p. 1

  • Method

    The survey categorizes communication techniques by network layer and performance target and reviews UAV-inclusive architectures and conventional optimization and machine-learning approaches.

    Source: §I · p. 1, C. Performance targets · p. 5

  • Results

    The survey concludes that UAVs provide an aerial layer in maritime architectures and support maritime IoT and broadband service use cases.

    Source: §IX · p. 15

  • Takeaways & Limitations

    The survey presents network architectures and open issues to stimulate further interest in UAV-aided maritime communications.

    Source: §I · p. 1

    Most related optimization studies evaluate limited network topologies to reduce computational complexity.

    Source: C. Machine learning · p. 14

Abstract

from arXiv · show

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

I. INTRODUCTION … C. Contributions

The paper surveys how UAVs can extend maritime connectivity beyond shore-based and satellite networks, organizing existing solutions by network architecture, layer, and performance target. It also reviews deployments and trajectory optimization and identifies open challenges for UAV-aided maritime communications.

I. INTRODUCTION

  • UAVs offer flexible, autonomous radio-resource provisioning for broadband, reliable, low-latency maritime coverage, addressing the limitations of land-focused networks and high-latency, low-data-rate satellite links.The survey argues that integrating UAVs is important for ubiquitous connectivity supporting maritime applications on the water surface and underwater.Source: §I · p. 1

A. Maritime communications

  • Maritime networks must serve heterogeneous vessels, buoys, platforms, vehicles, sensors, and actuators across applications with differing QoS needs, from broadband access to real-time video and URLLC.Use cases include trade, ocean exploration, pollution monitoring, marine tourism, and search and rescue.Source: A. Maritime communications · pp. 1–2
  • Existing shore-based base stations and satellite constellations cannot meet emerging maritime applications’ data-rate, delay, and reliability needs, while aerial nodes can reduce path loss and delay and add transmission paths.The survey notes that research on flexible and intelligent maritime networks remains necessary despite industrial broadband satellite and long-distance shore-to-vessel initiatives.Source: A. Maritime communications · p. 2

B. UAV-aided wireless networks

  • In maritime settings, UAVs can relay between ground base stations and vessels, support underwater-IoT data collection with USVs and UUVs, and provide LoS links for real-time SAR video.Energy-efficient communication, trajectory optimization, and wireless power transfer are relevant to extending UAV flight time.Source: B. UAV-aided wireless networks · p. 2

C. Contributions

  • The survey presents maritime network architectures incorporating UAVs, explains their roles and building blocks, and situates them alongside terrestrial, satellite, and heterogeneous maritime nodes.It frames UAV-aided solutions as a flexible complement to terrestrial and satellite segments with path-loss and delay reduction advantages.Source: C. Contributions · p. 2
  • The survey categorizes physical-layer, resource-management, and cloud/edge UAV-aided algorithms by performance target and thoroughly reviews trajectory optimization for maritime applications.These topics are presented as part of its overview of solutions addressing maritime network-layer issues.Source: C. Contributions · p. 2
  • The survey presents experimental studies of UAV-aided maritime communications and provides implementation details to clarify the status of real-world deployments.Source: C. Contributions · p. 2
  • The survey highlights open issues in IRS-aided communications, NOMA, swarm intelligence across UAVs, USVs, and UUVs, and extending UAV flight time through WPT.It aims to address a literature gap because prior surveys either included few UAV-aided maritime studies or omitted UAV-based solutions.Source: C. Contributions · p. 2

D. Structure

The survey progresses from UAV-aided maritime network architectures through technical solutions and experimental implementations to open issues and conclusions.

Section II: Architecture designs

  • Section II presents UAV-aided maritime network architecture designs and their components.The survey structure is also illustrated in Fig. 1.Source: D. Structure · pp. 2–3

Section III: Physical-layer issues

  • Section III reviews physical-layer issues in UAV-aided maritime communications.Source: D. Structure · p. 3

Section IV: Resource management

Section V: Cloud/edge computing and caching

  • Section V covers cloud/edge computing and caching to improve UAV-aided maritime application performance.Source: D. Structure · p. 3

Section VI: UAV trajectory design

  • Section VI addresses optimal UAV trajectory design in maritime environments.Source: D. Structure · p. 3

Section VII: Experimental implementations

  • Section VII discusses experimental implementations of UAV-aided maritime topologies.Source: D. Structure · p. 3

Section VIII: Open issues

Section IX: Conclusions

II. UAV-AIDED MARITIME NETWORK ARCHITECTURE … 4) Space segment:

The maritime communications network combines maritime, shore, aerial, and space segments to connect mobile maritime nodes. UAVs add flexible coverage and relaying, while satellites provide backup coverage and network backhaul or fronthaul.

A. MCN segments

  • The MCN architecture has four segments: maritime, shore, aerial, and space.The network links heterogeneous maritime nodes, UAVs, and satellites in mobile, volatile environments.Source: §II · p. 3, A. MCN segments · p. 4

1) Maritime segment:

  • Maritime edge nodes acquire data, relay cooperatively, and perform edge computing, while underwater sensor nodes, buoys, and UUVs use acoustic links for long-range exchange.Surface ships, USVs, and buoys support transportation, environmental observation, underwater relaying, and maritime search and rescue.Source: 1) Maritime segment · p. 4

2) Shore segment:

  • Shore base stations use cellular standards to cover nearby maritime nodes and UAVs, while ground stations connect to the space segment for broader MCN data transmission.Source: 2) Shore segment · p. 4

3) Aerial segment:

  • UAVs address broadband and URLLC coverage gaps by dynamically provisioning remote radio resources with lower latency than satellite links and reliable multi-hop, diverse wireless paths,.They also support maritime IoT relaying, search and rescue, intelligent transportation, and multimedia services; the aerial segment may include stratospheric high-altitude platforms.Source: 3) Aerial segment · p. 4

4) Space segment:

  • The space segment includes GEO systems such as INMARSAT and LEO constellations such as Starlink, providing backup coverage and backhauling or fronthauling across the MCN.Source: 4) Space segment · p. 4

B. Communication Technologies … 4) Acoustic communications:

Maritime communication networks use different technologies to accommodate heterogeneous segments and environmental conditions; most solutions are wireless, though some underwater topologies use wired links. The section surveys RF, optical, and acoustic wireless approaches and their capabilities and constraints.

B. Communication Technologies

  • Most maritime communication network solutions are wireless, while some underwater topologies rely on wired connections to accommodate heterogeneous segments and propagation conditions.Source: B. Communication Technologies · p. 4

1) Radio frequency (RF) communications:

  • Conventional maritime radio services and search-and-rescue support use very high frequency (VHF) transmissions in the 156–174 MHz band.Source: 1) Radio frequency (RF) communications · p. 5
  • Cellular and Wi-Fi bands support broadband and maritime IoT needs for high data rates and low latency, while UAV repositioning, direct terminal links –, and IRS can improve connectivity or mitigate fading path-loss.Source: 1) Radio frequency (RF) communications · p. 5

2) Free space optical (FSO) communications:

  • FSO offers large bandwidth, unlicensed spectrum, high data rates, fast deployment, and reduced power under LoS, but atmospheric absorption, scattering, and turbulence impair transmissions,.Hybrid RF/FSO schemes switch technologies as environmental conditions change and address misalignment, resource allocation, and energy efficiency –.Source: 2) Free space optical (FSO)… · p. 5

3) Visible-light communications (VLC):

  • VLC uses LEDs to avoid RF interference and signal leakage and can link underwater UUVs with one another and with USVs, ships, or UAVs acting as data sinks [41].Maritime VLC must account for deployment mobility and accurate transceiver pointing.Source: 3) Visible-light communications (VLC) · p. 5

4) Acoustic communications:

  • Acoustic waves enable long-range underwater transmission, but acoustic channels have poor quality, vary dynamically, offer less capacity than RF, and incur greater propagation delay –.Hybrid schemes combine acoustic signals with RF and optical signals to combat attenuation.Source: 4) Acoustic communications · p. 5

C. Performance targets … A. Channel modelling for maritime environments

The sections frame maritime-network design as balancing competing performance targets and implementation complexity, then survey UAV-aided physical-layer goals and channel models shaped by ocean conditions. They cover spectral efficiency, energy and lifetime, delay, offloading, and maritime-specific propagation.

C. Performance targets

  • Maritime-network procedures balance potentially conflicting performance targets against implementation complexity and can be optimized with conventional techniques and/or machine learning [47].Source: C. Performance targets · p. 5

1) Spectral efficiency:

  • Spectral efficiency, measured in bps/Hz, is pressured by growing node counts and network overlap, motivating frequency reuse through aerial base stations or other bands and intelligent link selection.Satellite-aided backhaul adds heterogeneity to maritime networks.Source: 1) Spectral efficiency · pp. 5–6

2) Energy efficiency/network lifetime maximization:

  • Energy-efficient communication, power control, and routing [49] support battery-dependent nodes and network lifetime, while wireless power transfer can prolong operation,.Energy efficiency is measured in bits/joule; sustainable operation also seeks to reduce carbon footprint and operating expenditure.Source: 2) Energy efficiency/network lifetime… · p. 6

3) Communication delay minimization:

  • For critical maritime services, end-to-end delay is a key QoS parameter; high-throughput route selection, channel quality, multiple antennas, and non-orthogonal, grant-free access can reduce it.Source: 3) Communication delay minimization · p. 6

4) Task and data offloading:

  • Edge computing can reduce latency when task allocation is efficient, while higher cache hit ratios improve QoS and backhaul/fronthaul offloading by avoiding remote-server data fetching.Cache hit ratio is the share of requested files among files stored in the cache.Source: 4) Task and data offloading · p. 6

III. PHYSICAL-LAYER DESIGN

  • UAV-aided maritime physical-layer design targets acceptable user QoS, 5G/6G channel models suited to harsh seas, higher data rates, and energy-efficient transmissions under limited UAV battery life.Source: §III · p. 6

A. Channel modelling for maritime environments

  • Ocean propagation differs from terrestrial channels because direct paths, sea volatility, and extreme weather matter; models address sea-surface shadowing and mobile LoS, sea-bounce, and waveguide paths.The FDTD results were limited to fixed transmitter and receiver heights and frequencies below 3 GHz; the model supports arbitrary movement of vessels, UAVs, and clusters. Figure 3 depicts the latter model.Source: A. Channel modelling for maritime… · pp. 6–7, Figure 3 · p. 7
  • Maritime channel studies use experimentally verified ray tracing despite omitting sea motion, decentralized ML for mmWave estimation, and describe long-range interference from atmospheric ducting.The ray-tracing study found received-signal amplitude followed a Rice distribution; the ML framework improved estimation accuracy and downlink data rates in simulation.Source: A. Channel modelling for maritime… · p. 7

B. Data rate maximization · C. Interference mitigation

The sections survey optimization and learning approaches for maritime UAV data-rate maximization and interference mitigation, covering link design, network coordination, and data acquisition. They also describe spectrum-sharing challenges and methods to manage co-channel interference.

B. Data rate maximization

  • [68] studies a UAV-enabled FSO wideband link to a supporting vessel with SAR capability; wind and snow together impair performance most, followed by fog, snow, and wind separately.The performance evaluation considers Arctic deployment and weather conditions.Source: B. Data rate maximization · p. 7
  • [59]’s normalized LMS distributed beamforming outperforms conventional LMS in convergence and spectral efficiency, with the SE gain more evident at high SNR.The method jointly optimizes the transceivers in a mmWave-based UAV-aided maritime communications network.Source: B. Data rate maximization · p. 7
  • [60] maximizes the minimum vessel rate through user-centric UAV/TBS clustering and edge control, while keeping satellite-user leakage interference below an acceptable threshold.The proposed optimization framework is reported to have improved convergence.Source: B. Data rate maximization · p. 7
  • [61] finds that underwater-link beamwidth should widen as transmission power increases to counter pointing error, but narrow when attenuation dominates signal deterioration.The study considers two-hop transfer using UAVs, UUVs, and a buoying relay, targeting low outage probability.Source: B. Data rate maximization · p. 7
  • [62] models and simulates connectivity from underwater sensors through surface sink relays to UAVs, and proposes edge-computing data acquisition evaluated across transmission factors.The sink nodes forward received acoustic signals to UAVs over wireless links; the connectivity model accounts for UAV trajectory, antenna characteristics, and sink stability.Source: B. Data rate maximization · p. 7

C. Interference mitigation

  • [22] notes that UAV/MCN co-channel interference is unstable under UAV motion and discusses sea-lane-based on-demand placement, maritime constraints, network coordination, satellite-backhaul limits, and trajectory-derived CCI patterns.When UAVs are far from shore, satellites may be the only backhaul option, with large delay and limited communication rate.Source: C. Interference mitigation · pp. 7–8
  • [63]’s two-step power-allocation algorithm, based only on large-scale CSI, converges quickly and improves spectral efficiency relative to other power-allocation schemes.Source: C. Interference mitigation · p. 8
  • [64]’s DRL approach trains neural networks for power allocation and message transmission, significantly improving BER and reducing overall power consumption.The two networks are trained for their respective tasks in UAV-aided maritime communications networks.Source: C. Interference mitigation · p. 8

D. Energy-efficient transmissions · IV. RADIO-RESOURCE MANAGEMENT · A. Cooperative UAV data relaying

The survey describes energy-aware UAV maritime networking and radio-resource management alongside cooperative relaying applications, from ocean monitoring to integrated space-air-sea networks. It highlights energy reduction, connectivity, coverage, throughput, delay, and network lifetime as design concerns.

D. Energy-efficient transmissions

  • In, an RSS-based probabilistic genetic handover model reduces unnecessary triggers using SNR, available rate, residual energy, and handover data.The same subsection describes ’s flexible-substrate cylindrical patch antenna sub-arrays, which adapt radiation patterns to applications, and notes radar-communication integration as a 6G consideration.Source: D. Energy-efficient transmissions · p. 8
  • In, large-scale-CSI link scheduling and rate adaptation improve network energy consumption over alternatives while converging quickly with performance similar to the optimal solution.The approach jointly schedules links and adapts rates in a hierarchical satellite-UAV-terrestrial network with QoS guarantees; Fig. 4 compares energy consumption across vessel occupied-time-slot cases.Source: Figure 4 · p. 8, D. Energy-efficient transmissions · p. 8
  • In, trajectory and in-flight transmit power are jointly optimized using large-scale CSI under UAV kinematics, interference, backhaul, and communication-energy constraints.Source: D. Energy-efficient transmissions · p. 8

IV. RADIO-RESOURCE MANAGEMENT

  • Radio-resource management targets spectral efficiency, effective UAV coverage, longer network lifetime to reduce outage, and authentication to address IoT security risks.The survey presents these as diverse design goals for UAV-aided maritime communications.Source: §IV · p. 8

A. Cooperative UAV data relaying

  • UAVs can serve as active base stations or relays; in oceanic monitoring networks, they relay between underwater sensors, surface sinks, and terrestrial base stations.The architecture has underwater transmission from underwater sensor nodes to sea-surface sink nodes, followed by an RF segment from sinks to ground stations or UAVs.Source: A. Cooperative UAV data relaying · p. 9
  • Cooperative relay studies address buoy throughput, vessel-aware delay and success, coastal container link quality, multi-UAV throughput, and UAV-USV task assignment.[70] reports improved minimum throughput for drifting buoys; reduces average delay and raises transmission success ratio; improves link quality over vessel-BS connections. achieves performance close to optimal for a scenario including 15 UAVs and 50 USVs, while uses matching to assign UAVs to USV coalitions.Source: A. Cooperative UAV data relaying · p. 9
  • In, restricted three-sided matching coordinates users, HAPs, and satellites; in, DRL-based node access improves data transfer rate, average D2D delay, and network throughput.These works consider integrated space-air-ground or terrestrial, sea, and high-altitude-platform communications.Source: A. Cooperative UAV data relaying · p. 9

B. Non-orthogonal multiple access … A. Mobile edge computing

The section surveys UAV-aided maritime communication approaches spanning NOMA, network lifetime, security, and edge computing. It describes how these approaches address connectivity, resource use, latency, energy, and authentication in maritime networks.

B. Non-orthogonal multiple access

  • NOMA shares resource blocks among nodes; maritime studies use cluster-based transmission and vessel groups served by UAVs or TBSs to improve spectral efficiency and reduce system cost.For, the reported ergodic sum-rate performance is superior to the OMA alternatives,; decomposes its non-convex problem into power-allocation and transmission-duration subproblems.Source: B. Non-orthogonal multiple access · p. 10

C. Network lifetime maximization

  • UAV relays between sensor nodes and TBSs support network-lifetime optimization,, with separate subproblems for UAV–sensor delay minimization and sensor-network lifetime maximization.The passage reports improved performance over other TDMA schemes. It frames lifetime maximization as a means to reduce outage probability and support high-data-rate end-to-end applications.Source: C. Network lifetime maximization · p. 10
  • An optimized link-state routing protocol using link expiration time and residual energy can improve end-to-end delay, packet transmission rate, and routing overhead.Source: C. Network lifetime maximization · p. 10

D. Security considerations

  • For 6G-IoT MCNs, a lightweight authentication protocol offers a better security-to-efficiency trade-off, while an elliptic-curve-based scheme improves computational and communication cost.The first approach is assessed using formal security methods; the second is evaluated experimentally.Source: D. Security considerations · p. 10

V. CLOUD/EDGE COMPUTING AND DATA OFFLOADING

  • MCN edge servers, sometimes mounted on UAVs, are deployed near ship terminals or maritime IoT to offload computing tasks; studies consider communication and energy use alongside latency and complexity constraints.The section introduces cloud/edge computing and data offloading, with key considerations outlined in Table V.Source: §V · p. 10, A. Mobile edge computing · p. 10

A. Mobile edge computing

  • For MEC-UAV network profit, an iterative Benders-decomposition algorithm offers near-optimal solutions with efficient convergence; multi-agent DRL for mobile data centers predicts user preferences to improve caching and reduce task latency and energy use.Source: A. Mobile edge computing · p. 11
  • UCB-based UAV edge-server selection accounts for budget, delay, and energy factors and lowers offloading latency and weighted latency-energy cost [97],; DRL improves resource utilization and reduces task implementation latency.Fig. 6 depicts a centralized, reward-based DRL architecture for optimizing a UAV-aided MCN.Source: A. Mobile edge computing · p. 11, Figure 6 · p. 11
  • For a two-layer maritime MEC network, deep Q-network and deep deterministic policy gradient algorithms optimize flight trajectories and virtual-machine participation to reduce total average latency; a joint Benders and primal decomposition approach yields near-optimal weighted sum-rate performance with efficient convergence.The same subsection also discusses an MEC-enabled space-air-ground-aqua architecture and its research challenges and future directions.Source: A. Mobile edge computing · p. 11

B. Data Offloading · VI. UAV TRAJECTORY OPTIMIZATION · A. Enhancing communication performance

The merged sections survey maritime UAV data-offloading and trajectory-optimization approaches, covering time-related offloading objectives and communication-performance goals such as improved connectivity, energy efficiency, and data rate. They describe optimization and learning methods for coordinating UAV placement, trajectories, scheduling, and resource use.

B. Data Offloading

  • [88] applies penalty convex-concave optimization to maritime data offloading and reports reductions in average offloading delay and offloading success ratio.The model accounts for container-vessel mobility, UAV movements, and wireless propagation over the sea.Source: B. Data Offloading · p. 11
  • [94] reports reduced overall computation-task execution time, while uses delayed deep deterministic learning to reduce total mission completion time.[94] studies incentive-based UAV–USV collaborative offloading; jointly considers UAV trajectory, offloading capability, buoy–UAV association, and transmit power.Source: B. Data Offloading · pp. 11–12

VI. UAV TRAJECTORY OPTIMIZATION

  • Trajectory optimization is linked to network lifetime, energy consumption, connectivity quality, and, in emergencies, network installation time.The section organizes related work around communication-performance enhancement and search-and-rescue procedures.Source: §VI · p. 12

A. Enhancing communication performance

  • [99] provides a Matlab/Simulink UAV–USV simulation platform, while and optimize maritime UAV placements for energy consumption and offshore-relay link capacity.[100] formulates integer linear programming solved with Gurobi; uses particle swarm optimization for a minimum-maximization link-capacity problem.Source: A. Enhancing communication performance · p. 12
  • [102]– jointly optimize buoy–UAV scheduling and flight trajectories under wind effects to reduce UAV energy consumption.Their successive convex approximation method uses a cyclical trajectory framework designed to handle arbitrary data volumes efficiently under wind effects.Source: A. Enhancing communication performance · p. 12
  • [104] uses successive convex approximation to jointly optimize power, transmission scheduling, and UAV trajectory for energy efficiency, with improved convergence time over other methods.[105] applies Fermat-point trajectory planning to maritime IoT with USVs, reporting improved data collection rate and fast convergence.Source: A. Enhancing communication performance · p. 12
  • [106]– address link quality, data rate, energy use, data collection, and multi-UAV task allocation through trajectory and placement optimization.[106] uses dual Q-learning to minimize average loss; combines power allocation and trajectory planning using large-scale CSI; derives single-user placement and evaluates multi-user placement; uses PSO across CSI scenarios and a Kalman filter for position errors; models random task allocation and single-UAV 2D route planning.Source: A. Enhancing communication performance · p. 12

B. Improving maritime search and rescue · VII. EXPERIMENTAL STUDIES

The sections survey UAV-aided maritime search-and-rescue approaches and experimental evaluations, spanning coordinated search, communications optimization, and field or laboratory demonstrations.

B. Improving maritime search and rescue

  • SAR research addresses UAV deployment and patrol scheduling, while emergency UAV base stations jointly optimize power allocation and trajectory to maximize throughput,.The emergency scenario broadcasts a node’s received signal to nearby users, and its optimization is solved with a loop iterative algorithm.Source: B. Improving maritime search and rescue · p. 12
  • Search studies link UAV altitude to rescue time and evaluate maritime target tracking and convoy-threat detection using UAV ad-hoc networks [112],.The computational tool considers collaborative search and tracking, as well as circular formation flight.Source: B. Improving maritime search and rescue · p. 13
  • A UAV–USV cooperative search path planned with reinforcement learning evaluates reward functions for improving system throughput.The study compares achievable throughput under different reward functions.Source: B. Improving maritime search and rescue · p. 13
  • An improved PSO coordinates UAVs, USVs, and UUVs across underwater-object search and tracking, while multiobjective PSO optimizes UAV radar-surveillance trajectories,.The first study seeks to maximize search space and minimize terminal error; the integrated system is reported more efficient than the USV–UUV system for these procedures.Source: B. Improving maritime search and rescue · p. 13

VII. EXPERIMENTAL STUDIES

  • Experiments demonstrate pressure-tolerant ocean-observing components at 200m depth and UAV sea-surface links to UUVs near shore with robust hovering,.UAVs and USVs support glider deployment, recovery, battery charging, and communication relay.Source: §VII · p. 13
  • A UAV–USV S-band campaign measures large- and small-scale channel characteristics, while BLUECOM+ relay positioning with PSO outperforms fixed or random height selection in throughput,.The BLUECOM+ height-control approach accounts for reflected signals to maximize reception quality; experiments also evaluate a UAV relay with directional antennas.Source: §VII · pp. 13–14
  • LTE-based SAR evaluations include a resource-guaranteed persistent-scheduling scheme that improves latency and reliability, multi-link experiments enabling smooth handovers, and reliable direct or relayed networks –.The studies include hardware implementation, software-defined-radio laboratory evaluation, large-scale experiments, and field trials involving UAVs, USVs, and UUVs.Source: §VII · p. 14

VIII. OPEN ISSUES · A. Physical-layer issues · B. UAV-aided non-orthogonal multiple access

The open issues span physical-layer designs and NOMA for maritime networks, alongside a need for 6G-frequency measurement campaigns. Proposed directions seek improved coverage, efficiency, reliability, and spectral efficiency while addressing deployment costs and delay tradeoffs.

VIII. OPEN ISSUES

  • UAV integration in maritime communication networks can improve coverage, delay, reliability, and deployment flexibility, but the research area is nascent and has open directions to explore.Source: §VIII · p. 14

A. Physical-layer issues

  • mMIMO can improve spatial separation, spectral efficiency, and energy efficiency, but UAV deployment raises installation costs and hardware complexity; d-MIMO is an active alternative.d-MIMO may reduce spatial correlation and path loss while increasing diversity and multiplexing gain.Source: A. Physical-layer issues · p. 14
  • UAVs carrying IRSs can enhance wireless channel quality by phase-shifting relayed signals,.Source: A. Physical-layer issues · p. 14
  • Measurement campaigns for 6G carrier frequencies remain an open direction; cited performance evaluations, including, have considered frequencies only up to 5 GHz.Source: A. Physical-layer issues · p. 14

B. UAV-aided non-orthogonal multiple access

C. Machine learning … IX. CONCLUSIONS

The survey presents UAVs as an aerial layer for maritime connectivity and reviews their architectures, optimization approaches, trajectory design, experiments, and open challenges for 6G. Its future directions emphasize scalable learning, secure and autonomous swarms, and edge services that support real-time maritime operations.

C. Machine learning

  • Realistic large-scale evaluation of maritime networks with active IoT devices is needed, and DRL can adapt optimization to changing conditions,.Edge solutions combined with federated learning are promising for MCNs in 6G; Fig. 7 depicts local model updates across K UAV-aided coverage areas.Source: C. Machine learning · p. 14, Figure 7 · p. 15

D. Security and safety

  • UAV identity-forging attacks can be difficult to detect in swarms, making authentication and received-content analysis important for communication reliability and infrastructure safety.Authentication and content analysis may affect delay performance.Source: D. Security and safety · pp. 14–15
  • Maritime UAV swarms face false-data injection and eavesdropping risks, motivating federated learning for privacy and distributed, low-complexity physical-layer security algorithms [140]–.Federated averaging aggregates weighted local model updates; physical-layer security approaches should exploit machine learning for autonomous operation.Source: D. Security and safety · p. 15

E. Advanced edge maritime services

  • UAV-aided mobile edge computing can process data locally to reduce backhaul use, while integrating UAVs, USVs, and UUVs can support task allocation and real-time collection.Edge computing and caching should jointly account for processing capacity, vehicle trajectories, and energy constraints to maintain UAV availability.Source: E. Advanced edge maritime services · p. 15

F. UAV swarm intelligence

  • UAV swarms need reliable inter-UAV connectivity and distributed intelligence to exchange trajectory and velocity data, avoid collisions, and improve coverage and task offloading [147].Bandit-based channel prediction can address outdated channel-state information, while wireless power transfer can mitigate energy limitations; delay, poor channel quality, harsh weather, and energy constraints remain challenges.Source: F. UAV swarm intelligence · p. 15

IX. CONCLUSIONS

  • The survey reviews UAV-aided maritime architectures, optimization and machine-learning solutions, trajectory design, experimental implementations, and open issues for 6G maritime activities.It frames UAVs as an intermediate aerial layer complementing shore- and satellite-based deployments.Source: §IX · p. 15

Related papers

Loading 2209.09605v1…