PaperlayerView exact revision on arXiv

Source-linked AI summary

An Extensive Survey on the Internet of Drones

Pietro Boccadoro, Domenico Striccoli, Luigi Alfredo Grieco

arXiv:2007.12611v3eess.SYeess.SP

TL;DR

IoD networks offer mobility, coverage extension, and access to otherwise unreachable locations, but face unreliable wireless links, energy constraints, mobility, security, and privacy challenges. This paper systematically classifies the literature across protocol layers and cross-layer approaches, extending the survey to applications, economics, open issues, and future directions.

  • Problem

    IoD research spans wireless unreliability, battery duration, mobility-driven topology changes, security, and privacy, while existing studies address the field’s themes in fragmented ways.

  • Method

    The paper studies and classifies IoD literature by protocol-stack layer and finer layer-specific issues, including cross-layer approaches, applications, economics, security, privacy, and emerging technologies.

  • Results

    The analysis identifies layer-specific design efforts for communication range, wireless-medium modeling, routing, synchronization, cooperation, and network formation, while characterizing open research issues.

  • Takeaways & Limitations

    The survey provides a structured view of IoD research trends and highlights research directions spanning connectivity, resource allocation, coordination, security, privacy, and economics.

  • Takeaways & Limitations

    IoD adoption remains constrained by unresolved privacy concerns and by drone cost and accessibility, particularly for industrial applications.

Abstract

from arXiv · show

The Internet of Drones (IoD) recently gained momentum due to its high adaptability to a wide variety of complex scenarios. Indeed, Unmanned Aerial Vehicles (UAVs) can successfully be employed in different applications, thanks to some technological and practical advantages: high mobility, capability to extend wireless coverage areas, or ability to reach places inaccessible to humans. Moreover, the employment of drones promisingly improves the performance parameters of different network architectures. Nevertheless, the adoption of networks of drones gives rise to several issues related to the unreliability of the wireless medium, the duration of batteries, and the high mobility degree, which may cause frequent topology changes. Also security and privacy issues need to be properly investigated. With respect to other surveys on IoD-related topics, the goal of the present work is to categorize the multifaceted aspects of IoD, proposing a classification approach of the IoD environment that develops along two main directions. At a macroscopic level, it follows the structure of the Internet protocol stack, starting from the physical layer and extending to the upper layers, without neglecting cross-layer and optimization approaches. At a finer level, all the most relevant works belonging to each layer of the stack are further classified, according to the different issues peculiar of the layer, and highlighting the most relevant differences with the other surveys present in literature. To provide a deeper insight in the theme, the present work embraces many facets of the IoD, including privacy and security considerations as well as the potential economic impact of the IoD. Finally, a discussion on the main research challenges and possible future directions is carried out, focusing on the open issues and the most promising technologies that deserve to be further developed in the IoD field.

1. Introduction

The paper surveys IoD research as a network architecture connecting autonomous flying vehicles with ground entities, while addressing wireless, energy, mobility, security, and privacy challenges. It organizes this multifaceted literature across protocol layers, cross-layer approaches, applications, economics, and future directions.

  • IoD connects autonomous flying vehicles with ground network entities that coordinate drones, process data, control airspace, and support remote services.
  • Unreliable wireless links and high drone mobility cause data losses, interruptions, limited communication range, topology changes, and coordination challenges.
  • Drone energy use includes sensing, communication, computation, movement, flight patterns, and mission plans, while high traffic volumes also make onboard memory relevant.
  • Its classification follows the Internet protocol stack from physical through application layers, separately treating cross-layer and optimization approaches before discussing research challenges and future directions.
  • The survey addresses IoD security and privacy while examining emerging technologies including mmWave, VLC, 6G, and ICN, and discusses drone economics.

2. Related Survey and Review Papers

Existing IoD surveys examine important but often specialized topics, including communication technologies, routing, energy efficiency, channel modeling, coverage, optimization, and mmWave. This paper responds with a broader layer-by-layer taxonomy spanning technological aspects and research questions.

  • Earlier surveys focus on specific domains such as IoT communication technologies, routing and energy efficiency, or legislation and security in UAV-assisted cellular networks.
  • The reviewed literature is fragmented across topics, with limited wide-range analysis of challenges spanning different protocol-stack layers.
  • Related works separately study cooperation models, physical-layer channel modeling, surveillance robotics, UAV positioning optimization, area coverage, and mmWave communications.
  • Its organization includes IoD application fields and economics alongside physical-, data-link-, network-, transport-, and application-layer themes.
  • The proposed survey uses a cross-cutting, layer-by-layer classification of mathematical formulations, technologies, and implementation proposals.

3. IoD applicability

This section surveys IoD applicability across operational contexts, reference architectures, application fields, and potential economic effects. It presents drones as versatile networked systems spanning telecommunications, disaster response, agriculture, industry, government, and logistics.

  • 3. IoD Application Fields: IoD applicability is organized through a taxonomy that examines application fields and the economic perspective of drone adoption.The section uses Figure 3 to organize its analysis and Table 2 to summarize major drone applications.
  • 3.1. IoD Application Fields: Drones are considered for emergency connectivity, disaster response, environmental monitoring, agriculture, surveillance, transport, logistics, and other civil and industrial applications.The surveyed application fields include public protection, health, construction, manufacturing, telecommunications, transport, agriculture, mining, and energy.
  • 3.1. IoD Application Fields: IoD architecture is presented as general-purpose and versatile, supporting multiple requirements across application fields and drone roles.The reference architecture includes UAV swarms, ground control infrastructure, base stations, satellites, web applications, and cloud computing.
  • 3.2. IoD Economics: Drone-enabled delivery is associated with estimated improvements of ∼50% in transportation speed, almost 50% lower environmental impact, ∼40% greater package-flow control, and ∼30% improved safety.These estimates concern systems and methods for delivering mail and goods using UAVs.
  • 3.2. IoD Economics: The section concludes that widespread drone employment could automate and optimize processes while changing business process design and time-to-market.The passage presents this as a broad forecast whose precise impact remains difficult to foresee.

4. Physical Layer

The Physical Layer survey covers modulation, connectivity, throughput, channel modeling, and communication technologies. It organizes these topics as the principal technological and research challenges of IoD PHY design.

  • 4. Physical Layer: The Physical Layer section surveys modulation techniques, connectivity, throughput maximization, channel modeling and characterization, and communication technologies.Figure 5 presents the organization of the physical-layer taxonomy.

4.1. Modulation Techniques

This section reviews modulation techniques for UAV communications, emphasizing data-link efficiency, long-range communication, battery life, and electromagnetic-noise suppression. It covers SC-FDM and OFDM-based approaches.

  • 4.1. Modulation Techniques: The surveyed modulation literature includes contributions focused specifically on UAV communications.The section identifies works as its modulation-technique coverage.
  • 4.1.1. Signal Quality Enhancement: SC-FDM is studied for UAV links by comparing it with OFDM in transmission-power optimization, bit error rate, and pulse shaping.The stated goal is to support long-range communication and long battery life for UAV networks.
  • 4.1.1. Signal Quality Enhancement: An OFDM noise-suppression method filters electromagnetic impulse noise at the receiver before OFDM demodulation.The approach is analyzed using models of electromagnetic pulses and their influence on OFDM-based communication.

4.2. Connectivity

IoD connectivity research addresses coverage, resource allocation, relaying, routing, cooperation, path planning, position optimization, and application-specific networking. These studies model LoS and NLoS conditions across Air-to-Air and Air-to-Ground links.

  • 4.2. Connectivity: Connectivity studies span coverage and resource allocation, relaying and routing, cooperative networks, path planning and position optimization, and application-specific scenarios.The survey identifies these as the main connectivity research directions.
  • 4.2.1. Coverage Analysis and Optimization: Coverage analyses model UAV-to-ground service using directional antennas, LoS and NLoS components, and dominant-interferer effects.Figure 6 illustrates NLoS communications enabled by drones.
  • 4.2.2. Resource Allocation and Optimization: Connectivity-aware resource allocation models Air-to-Air and Air-to-Ground links using environmental, building-density, placement, and relative-position factors.One cited approach uses these models to allocate spectrum in a cache-enabled UAV network while addressing queue stability and bandwidth requirements.
  • 4.2.3. Relaying and Routing: Relaying and routing studies use connectivity to increase mmWave communication range, minimize latency, maximize Packet Delivery Ratio, and model blocked 5G mmWave backhaul links.The surveyed methods include mobile relaying, connectivity-based routing, and analytical backhaul-channel models.
  • 4.2.4. Cooperative Networks: Cooperative-network studies optimize inter-drone distance and continuity of connectivity for stable wireless data transfer and drone communication.The cited work includes distance control and drone infrastructure for exchanging data with ground users.
  • 4.2.5. Path Planning and Position Optimization: Path-planning and position-optimization studies use connectivity models to maximize coverage or connectivity while reducing interference and transmission delay.Some approaches also account for the time UAVs need to reach destinations.
  • 4.2.6. Connectivity in Specific Applications: A distinct connectivity approach optimizes user QoE based on data rate, delay, device type, and LoS-derived connection-loss probability.This links connectivity modeling directly to user-level service quality.

4.3. Throughput Maximization and Evaluation

The survey organizes throughput research across physical, link, and network layers, emphasizing UAV positioning, channel-aware optimization, access control, relaying, routing, and network formation.

  • Scope: The surveyed approaches span direct throughput optimization, link-quality evaluation, and cross-layer designs across heterogeneous IoD scenarios.The section explicitly covers physical, data-link, and network-layer strategies rather than a single protocol layer.
  • Physical layer: Throughput studies primarily target the physical layer, especially UAV position and path optimization for improving rates, latency, interference, or coverage.Several approaches jointly optimize UAV trajectories, transmission power, resource blocks, or placement using channel and network conditions.
  • Data link layer: Link-layer proposals improve throughput through VLC analysis, adaptive LTE-A framing, dynamic access control, resource allocation, and energy-aware frame selection.The surveyed mechanisms include one- and two-hop VLC configurations, relay scheduling, contention-window adjustment, and priority-based transmission.
  • Network layer: Network-layer work maximizes throughput through relay placement, energy-aware scheduling, throughput-oriented routing, and utility-based flying backhaul formation.Relay schemes balance outage probability, hop and cluster placement, energy consumption, bit errors, data rate, relayed packets, and delay.

4.4. Channel Modeling and Characterization

Channel-modeling research characterizes UAV communication links to support position, power, coverage, rate, network-formation, and equalization decisions across diverse environments and frequencies.

  • Position and energy optimization: Channel models are used to optimize UAV position and energy, including joint placement and power allocation for air-to-ground, ground-to-ground, and cognitive-radio systems.The surveyed studies derive path loss and channel gain or use channel models to determine UAV locations and operating power.
  • Performance analysis: Air-to-ground models support coverage, rate, SINR, SNR, load balancing, and network-formation analyses using line-of-sight and non-line-of-sight propagation.These models are applied to derive end-to-end rates and balance loads between UAV base stations and ground Wi-Fi access points.
  • Altitude and measurement: Altitude-focused studies examine trade-offs among power, capacity gain, path loss, scattering, and measured propagation characteristics.Measurement campaigns capture path loss, fading, power-delay profiles, multipath components, and delay spread at low UAV heights.
  • Specialized channel conditions: Additional models address mmWave propagation, obstacle-aware environments, Doppler and inter-carrier interference, and doubly selective channels.These efforts support clustering, user association, channel learning, frequency-domain estimation, tracking, and equalization.

4.5. Communication Technologies

The survey covers communication technologies selected according to UAV tasks, mobility, environment, range, throughput, and energy constraints, including Wi-Fi, mmWave, cellular, machine-type, cognitive-radio, and optical systems.

  • Wi-Fi: Wi-Fi supports ad-hoc drone networking, high-quality video transmission, distance control, service continuity, and applications requiring varied throughput.Its use is motivated by ease of deployment and unlicensed-spectrum access, including replacement of UAVs limited by battery lifetime.
  • mmWave: mmWave research addresses relay positioning, multiple access, beamforming, cellular UAV networks, backhaul blockage, channel characterization, and mobility-related tracking.The surveyed advantages include high frequency, increased bandwidth, and beamforming, while attenuation and blockage remain relevant design issues.
  • MTC and cognitive radio: Machine-type and cognitive-radio technologies extend IoD communication options through D2D discovery, clustering, relaying, opportunistic spectrum access, and spectrum sensing.These technologies are discussed for public-safety and UAV-based applications involving constrained or shared spectrum.
  • Other technologies: LTE-U and free-space optical systems support emergency coverage restoration and high-data-rate relay communication, respectively.LTE-U fills Wi-Fi coverage gaps through hybrid infrastructure, while buffer-aided mobile UAV relays are studied for FSO links.

4.6. Comparison with Other Surveys

Compared with prior surveys, this work broadens and structures IoD coverage by extending existing topic analyses and adding optimization, cross-layer, and less-covered application dimensions.

  • Prior survey scope: Prior surveys discuss physical-layer connectivity, throughput, channel modeling, and communication technologies, but often within specific scenarios, strategies, or technologies.The cited comparison identifies coverage-focused connectivity reviews, routing-oriented throughput discussions, and narrowly scoped channel studies.
  • Added coverage: This survey extends connectivity coverage to resource allocation, relaying, routing, cooperative networks, and UAV path planning and optimization.Its throughput treatment also reaches network-layer optimization while covering physical- and link-layer aspects.
  • Comparative contribution: The survey adds broader channel analysis linked to position and altitude optimization and organizes the literature across multiple IoD dimensions.The stated value is to enhance previously covered topics while adding topics absent from earlier surveys.

4.7. Lessons Learnt

IoD research links connectivity, channel modeling, throughput, and communication technology choices to coverage, resource use, reliability, and system performance, while exposing accuracy, computational, and energy trade-offs.

  • Modulation methods used in UAV networks are generally established techniques whose novelty lies mainly in their application to FANETs.LTE-derived techniques can improve noise reduction and transmission-power optimization, while modulated-signal detection supports drone detection and classification.
  • Connectivity and channel models support coverage, resource allocation, path reliability, routing, path planning, and drone-position optimization, but often rely on simplifying hypotheses.They may omit jointly modeling obstacles, signal dispersion, interference, drone height, and mobility.
  • Throughput optimization can improve data exchange and optimize drone positioning, paths, spectrum efficiency, energy efficiency, and network topologies.These gains commonly require computationally expensive, approximate, or simplified optimization solutions.
  • Higher throughput can require higher energy consumption, creating a serious drawback for battery-powered drones.
  • Communication technology selection should reflect environmental conditions, task type and duration, communication range, and drone energy consumption.Reviewed studies exploit technology-specific characteristics to optimize transmission aspects.

5. Data Link Layer

The data link layer survey organizes IoD work around resource optimization, scheduling, coding, and related MAC strategies. These approaches improve communication efficiency and packet delivery but introduce interoperability, overhead, synchronization, delay, and complexity costs.

  • 5.1. Resource Allocation: Resource-allocation studies optimize PDR, packet-error probability, bandwidth, delay, throughput, and energy efficiency.
  • 5.3. Coding Strategies: Network coding saves time and bandwidth by adding repair information, often with cooperative multi-hop techniques in UAV networks.Coded packets can improve spectral efficiency and QoS, including for video streams over dynamic topologies.
  • 5.2. Data Scheduling: Scheduling methods target successful data rate, throughput, energy consumption, and interference among drones.
  • 5.4. Comparison with Other Surveys: Link-layer research integrates resource optimization, scheduling, and coding strategies to improve UAV communication links across diverse operating conditions.The survey extends earlier analyses across a broader range of link-layer topics.
  • 5.5. Lessons Learnt: Link-layer improvements can require MAC modifications that create interoperability problems, while cooperative schemes add information overhead and complicate scheduling and synchronization.
  • 5.5. Lessons Learnt: Coding improves time, bandwidth, and spectral efficiency but increases decoding delay and encoding-decoding complexity compared with simple retransmission.This motivates research into low-complexity techniques that scale to large UAV networks.

6. Network Layer

The network-layer literature addresses cooperation, routing, and relaying in multi-hop UAV networks, targeting connectivity, latency, energy efficiency, coverage, and throughput. Across these approaches, effectiveness is strongly context-dependent, while joint optimization often requires high computational effort.

  • Cooperation: Cooperative strategies target connectivity, energy efficiency, data collection, coverage, and throughput across content delivery, IoT, sensor, cellular, and remote-sensing scenarios.Examples include moving content routers, low-delay cooperative routing, UAV-assisted data collection, cached-content clusters, and swarm relaying.
  • Relaying: A jointly optimized UAV relaying trajectory, speed, and time allocation reveals a trade-off between propulsion energy and overall spectrum and energy efficiency.The optimization concerns a circular relay trajectory and balances UAV propulsion consumption against system efficiency.
  • Lessons Learnt: Cooperation, routing, and relaying improve connectivity, routing effectiveness, energy efficiency, and latency in multi-hop UAV networks, but optimization is computationally costly.The surveyed works frame many of these improvements as optimization problems whose optimal solutions require substantial computational effort.
  • Routing: Routing strategies address UAV mobility, rapidly changing topology, and error-prone channels through proactive, predictive, centralized, analytical, and information-based approaches.Examples include reinforcement-learning routing, predictive connection-time estimation, dynamic drone replacement, and centralized ground-control architectures.
  • Relaying: Relay schemes extend coverage, reduce interference, improve energy efficiency and throughput, and defend against jamming by optimizing relay numbers and positions.Their benefits are especially relevant in environments affected by physical obstacles, external jammers, or short-range technologies such as mmWave.

7. Transport and Application Layer

Transport and application-layer studies examine QoE, offloading, video streaming, data handling, monitoring, and task allocation for heterogeneous IoD applications. The surveyed results show that mobility, repositioning, and application-aware optimization can support quality, coverage, resource use, and mission responsiveness.

  • Scope: Application studies cover QoE, computation offloading, video streaming, data collection and distribution, event monitoring and management, and task allocation.The section organizes these contributions within the transport and application-layer taxonomy.
  • Quality of Experience: Application requirements vary between very low latency for real-time monitoring and ultra-high quality for image and video transmission, motivating QoE indexes.The survey links these heterogeneous requirements to application-specific quality assessment.
  • Quality of Experience: QoE-aware flight planning improves users’ overall QoE, while adaptive fuzzy-logic routing can improve both QoS and QoE in infrastructure-limited FANETs.The cited approaches use Q-learning for flight-plan diversification and fuzzy logic for adaptive routing.
  • Computation Offloading: Offloading strategies address constrained onboard computation through MEC, response-time prediction, and MPTCP-based task-offloading decisions.These approaches target complex, time-consuming mission calculations and limited UAV resources, including hazardous-location surveillance.
  • Lessons Learnt: Optimizing paths, resource handling, and data exchange affects latency, packet delivery, QoE, onboard resource use, and mission longevity.The survey connects QoS indicators such as latency and PDR with QoE, especially for streaming services.
  • Lessons Learnt: Application-layer mobility enables temporary offloading and repositioning of mobile base stations or swarms, including redeployment when replacement is necessary.The survey identifies mobility as a key enabler across several application scenarios.

8. Cross-Layer and Optimization Approaches

Cross-layer and optimization studies jointly address path design, state estimation, network formation, control, and coverage while balancing energy, connectivity, throughput, delay, and mission objectives. Their central challenge is that these metrics conflict, mobility requires continual state updates, and tractable algorithms may simplify real scenarios.

  • Lessons Learnt: Energy minimization can increase flight time, battery duration, and delivered data, making it a major driver of cross-layer optimization.The survey links energy-focused optimization to trajectory design, state estimation, and resource handling.
  • Path Optimization: Trajectory studies optimize energy footprints, propulsion costs, throughput, speed, payload, and no-fly-zone constraints through theoretical models and convex or multi-variable optimization.The surveyed settings include fixed-altitude flight, hover-and-fly trajectories, truck-drone delivery, and MEC-supported uploading and downloading.
  • State Estimation and Optimization: Multi-objective swarm optimization can jointly plan paths and tasks while prioritizing coverage or connectivity according to mission commitments.A genetic-algorithm approach minimizes mission completion time while incorporating area coverage and communication paths.
  • Lessons Learnt: Cross-layer optimization targets energy consumption, connectivity, throughput, delay, task completion, mission time, and coverage across path, state, and network-control problems.The surveyed approaches span trajectory design, swarm task allocation, MEC-assisted communication, network formation, and coverage enhancement.
  • Lessons Learnt: Conflicting metrics require trade-offs, while multi-objective algorithms increase computational complexity and simplifications risk oversimplifying real application scenarios.High mobility also forces continuous state updates, increasing control-information overhead between UAVs and ground stations.

9. Security and Privacy Aspects

The IoD security and privacy literature spans communications security, authentication, threat assessment, application-specific risks, and privacy preservation, while also exposing regulatory gaps.

  • Security: Security analyses cover threats across drone-network components and assess risks in terms of integrity, confidentiality, and availability.The surveyed work also discusses future directions for securing unmanned systems.
  • Application Contexts: Security and privacy risks are examined across public-safety, smart-farming, military, and commercial applications, including domestic-use and ethical concerns.The literature also notes that differing national laws and insufficient legislation constrain civil and military drone governance.
  • Security: Authentication research addresses resource-constrained UAV networks through mechanisms including mutual UAV–ground-station authentication and secure communication protocols.Some proposed mechanisms combine authentication, key agreement, non-repudiation, and revocation while targeting energy efficiency.
  • Security: Wireless drone systems remain vulnerable to practical attacks, including effective deauthentication attacks against consumer WiFi drones without specialized hardware.The cited assessment uses a software suite rather than additional equipment such as software-defined radio.
  • Privacy Preservation: Privacy concerns arise because drones operate across urban and suburban areas, potentially impairing people’s privacy, especially with amateur deployments.A reviewed framework uses cognitive IoT for sensing, analytics, knowledge discovery, and intelligent decision-making in drone surveillance.

10. Discussion on the Main Findings and Future Research Perspectives

The surveyed literature shows that IoD performance depends on cross-layer connectivity, cooperation, optimization, and application design, while major open challenges remain in realistic operating conditions, security, privacy, standardization, and regulation.

  • 10.1. Discussion on the Main Findings in the Surveyed Literature: IoD connectivity research spans A2A and A2G technologies, but channel and connectivity models remain limited by NLoS conditions and obstacles.Selecting communication technology according to application needs is identified as a promising response.
  • 10.1. Discussion on the Main Findings in the Surveyed Literature: MAC-layer modifications, cooperative schemes, synchronization, and scheduling are central strategies for improving efficiency and resource allocation, especially in dense drone swarms.Scheduling is highlighted for densely populated scenarios.
  • 10.1. Discussion on the Main Findings in the Surveyed Literature: Drones can serve as relays to strengthen existing network solutions, while cooperative multi-drone applications support monitoring, surveillance, and rapid inspection of difficult-access areas.The surveyed applications associate quick on-site deployment with benefits for inspection systems.
  • 10.1. Discussion on the Main Findings in the Surveyed Literature: Throughput optimization addresses communication, positioning, routing, spectrum, energy, and topology objectives, but often requires computationally expensive optimization problems.The analysis covers both A2A and A2G data exchange.
  • 10.2. Research Challenges and Possible Future Directions: Future IoD development includes standardized heterogeneous 5G frameworks, Blockchain and smart contracts, VLC, ICN, and stronger security and privacy research.ICN is presented as relevant to mobility and handover, while privacy research remains limited and legislation is unresolved.
  • 10.3. Other Remarks: Industrial diffusion depends on technology costs, accessibility, and legal conditions because non-unified legislation can create economic and legal barriers to drone adoption.The paper connects widespread autonomous-vehicle deployment with industrial applications and cost considerations.

11. Conclusions

The survey organizes IoD research across protocol-stack layers and cross-layer approaches, identifies open issues, and reviews optimization and emerging communication technologies.

  • The survey classifies IoD literature by Internet protocol-stack layer and further groups studies according to their proposed approaches.
  • The survey identifies open issues and future research directions while reporting substantial design efforts across different IoD layers.
  • Combined optimization approaches address multiple IoD problems, including energy consumption with trajectory design and path planning with routing optimization.
  • The reviewed literature examines how 5G- and 6G-compliant technologies, including mmWave and VLC, affect IoD design.
Loading 2007.12611v3…