Unmanned Aerial Vehicles: A Survey on Civil Applications and Key Research Challenges

Hazim Shakhatreh; Ahmad Sawalmeh; Ala Al-Fuqaha; Zuochao Dou; Eyad Almaita; Issa Khalil; Noor Shamsiah Othman; Abdallah Khreishah; Mohsen Guizani

Source-linked AI summary

Unmanned Aerial Vehicles: A Survey on Civil Applications and Key Research Challenges

Hazim Shakhatreh, Ahmad Sawalmeh, Ala Al-Fuqaha, Zuochao Dou, Eyad Almaita, Issa Khalil, Noor Shamsiah Othman, Abdallah Khreishah, Mohsen Guizani

arXiv:1805.00881v1Apr 19, 2018cs.RO

TL;DR

Civil UAV use is expanding, but applications face unresolved operational and technical challenges. This survey synthesizes civil applications, research trends, and future directions, identifying energy, collision-avoidance, swarming, networking, and security challenges. It concludes that UAVs offer practical benefits across domains while highlighting open challenges for wider adoption.

Problem
Civil UAV applications span many domains, but unresolved challenges in energy, collision avoidance, swarming, networking, and security remain important research questions.
Method
The paper surveys civil UAV applications, reviews recent literature and research trends, and discusses future uses and approaches to open challenges.
Results
The survey identifies civil infrastructure as a major UAV application area and reports benefits including reduced injuries, inspection costs, and inspection time.
Takeaways & Limitations
UAVs can support monitoring, inspection, remote sensing, and other civil applications where human access is difficult or risky.
Takeaways & Limitations
UAV operations remain constrained by battery power, creating tradeoffs between real-time onboard analysis and storing data for later processing.

Abstract

from arXiv · show

The use of unmanned aerial vehicles (UAVs) is growing rapidly across many civil application domains including real-time monitoring, providing wireless coverage, remote sensing, search and rescue, delivery of goods, security and surveillance, precision agriculture, and civil infrastructure inspection. Smart UAVs are the next big revolution in UAV technology promising to provide new opportunities in different applications, especially in civil infrastructure in terms of reduced risks and lower cost. Civil infrastructure is expected to dominate the more that $45 Billion market value of UAV usage. In this survey, we present UAV civil applications and their challenges. We also discuss current research trends and provide future insights for potential UAV uses. Furthermore, we present the key challenges for UAV civil applications, including: charging challenges, collision avoidance and swarming challenges, and networking and security related challenges. Based on our review of the recent literature, we discuss open research challenges and draw high-level insights on how these challenges might be approached.

I. INTRODUCTION … A. UAV-Based SAR System

The survey frames UAVs as versatile, increasingly important civil-service platforms, while emphasizing application-specific and cross-domain challenges. It reviews UAV markets, classifications, civil applications, and SAR systems, including single- and multi-UAV operational workflows.

I. INTRODUCTION: UAVs support diverse civil applications because they are easy to deploy, inexpensive to maintain, highly mobile, and able to hover.Applications include traffic monitoring, wireless coverage, remote sensing, search and rescue, goods delivery, surveillance, precision agriculture, and infrastructure inspection.
I. INTRODUCTION: The survey addresses a gap in prior work by reviewing civil applications, application-specific challenges, research trends, and future insights.It also examines cross-domain challenges and practical ways to overcome challenges affecting multiple application domains.
II. MARKET OPPORTUNITY: $127 billion is the addressable market value of UAV uses, with civil infrastructure expected to dominate at $45 billion.The market opportunity includes equipment manufacturers, investors, and business service providers.
II. MARKET OPPORTUNITY: Smart UAVs may help manufacturers use new technological trends to overcome current application challenges, while global deployment requires a complete legal framework and regulatory institutions.The passage presents UAV applications as economically important for the near future.
III. UAV CLASSIFICATION: UAV classification must account for application-dependent features, including operating platform, with aerial communications platforms categorized as LAP or HAP.LAP operates below 10 km, whereas HAP operates above 10 km and can remain in the stratosphere for long periods.
IV. SEARCH AND RESCUE (SAR): UAVs are especially valuable for public safety, search and rescue, and disaster management involving floods, tsunamis, terrorist attacks, and critical infrastructure.The cited passage identifies water and power utilities, transportation, and telecommunications systems among affected infrastructure.
A. UAV-Based SAR System: UAV-based SAR reduces costs, resources, and human risks compared with traditional aircraft and helicopters, which require special training and permits.Traditional aerial SAR operations are described as typically very costly and as wasting substantial money and time each year.
A. UAV-Based SAR System: Single-UAV SAR scans a defined region with vision or thermal cameras and sends real-time imagery to the GCS, while multi-UAV SAR assigns planned trajectories for coordinated scanning and transmission.In multi-UAV systems, on-board imaging sensors locate missing persons, and target locations with related videos and images are transmitted to the GCS.

B. How SAR Operations Utilize UAVs … 3) Future Insights:

UAVs support SAR through aerial imaging, autonomous target search, image processing, and machine learning, while legislation, weather, energy, processing, coordination, localization, and communications remain key constraints and research priorities. Future work emphasizes sensor fusion, efficient onboard intelligence, long-duration operation, autonomy, swarm coordination, precise mapping, and emergency connectivity.

B. How SAR Operations Utilize UAVs: UAVs support SAR by surveying disaster areas with high-resolution imagery, evaluating infrastructure damage, and enabling autonomous, accurate searches without additional risks.A GPS-equipped lightweight quadrotor prototype was developed to help find lost persons.
1) Legislation:: U.S. FAA regulations currently prohibit autonomous UAV swarms for commercial use, although adjusted rules could enable swarms to coordinate SAR teams.The passage identifies regulatory adjustment as a possible pathway for commercial swarm use.
2) Weather:: Weather can divert UAVs from predetermined paths and cause mission failure during tsunamis, hurricanes, terrorist attacks, and other disasters.The passage characterizes adverse weather as a major challenge in such scenarios.
3) Energy Limitations:: Energy consumption is a major SAR challenge because batteries power hovering, communications, processing, and image analysis during extended disaster-region operations.Operators must decide whether to analyze data onboard in real time or store it for later analysis.
1) Image Processing:: Image processing enables rapid target detection in single- and multi-UAV SAR systems, using onboard or ground-station processing and location information augmented on aerial images.UAVs can integrate thermal and vision cameras, while terrestrial networks can transmit images and GPS locations to the ground control station.
2) Machine Learning:: Machine learning can identify frames containing lost persons, but onboard power and processing limits, adversarial attacks, and reliable real-time GCS communications remain challenges.The cited approach combines a pre-trained CNN with a trained linear SVM.
3) Future Insights:: Future SAR research should develop sensor-fusion algorithms, low-power onboard deep learning, power-efficient distributed swarm-data processing, and improved materials, batteries, and energy harvesting.These directions address detection accuracy, onboard resource constraints, real-time swarm data, and long-duration missions.
3) Future Insights:: Additional priorities include autonomous routing and collision avoidance, multi-hop swarm QoS, precise sensor-fused localization, and UAV aerial base stations for disrupted or overloaded networks.GPS coverage and accuracy issues motivate improved localization and mapping, while aerial base-station use remains nascent.

V. REMOTE SENSING … 3) Free Space Optical:

UAV remote sensing supports environmental monitoring, disaster management, atmospheric studies, and diverse data-driven applications through planned sensing, image processing, and emerging machine-learning, cloud, and FSO technologies. However, camera limitations, changing illumination, and challenging environments remain important barriers to reliable remote-sensing operations.

V. REMOTE SENSING: UAVs collect ground-sensor data, support environmental monitoring and disaster management, and provide datasets for crop, drought, water-quality, tree-species, and disease applications.These uses make UAVs both aerial sensor networks and sources of remote-sensing datasets.
A. Remote Sensing Systems: Active remote sensing transmits radiation toward an object and measures the reflected radiation, whereas passive systems rely on externally available energy.The passage identifies active and passive systems as the two primary remote-sensing types.
B. Image Processing and Analysis: UAV image processing uses autopilot logs for initial image positions and orientations, then aerial triangulation, automated tie points, and bundle-block adjustment to optimize image geometry.The workflow recovers accurate surface-point positions from image measurements and redundant observations.
C. Flight Planning: Flight planning defines the survey area from a background map or satellite image and incorporates altitude, camera focal length and orientation, flight path, and camera-trigger timing.The resulting information supports initial estimates used to recover exact surface-point positions.
1) Hostile Natural Environment:: UAVs access hazardous environments for atmospheric, air-quality, climate, and ocean studies, with rotary-wing aircraft suited to hovering and fixed-wing aircraft suited to longer distances and higher altitudes.An Aerosonde reconnaissance flight into Typhoon Longwang lasted 10 hours and was divided into four flight legs.
2) Camera Issues:: Lightweight UAV cameras remain constrained by radiometric and geometric limitations, bulky commercial instruments, calibration concerns, and spectral drawbacks for remote-sensing applications.The passage notes that current general-market cameras are not optimized for remote sensing.
3) Illumination Issues:: Strong shadows and rapidly moving clouds can disrupt automated image matching, cause aerial triangulation failures, and introduce errors into automatically generated digital elevation models.Changing shaded areas between images are especially problematic during the same mission.
1) Machine Learning:: Remote-sensing machine learning proceeds through UAV data collection, data cleansing, variable reduction when needed, and selection of an algorithm suited to the problem.Cloud GIS can combine UAV-derived NDVI maps with visualization and decision-support tools, while FSO can transmit large image and video volumes to command centers via satellite communication.

4) Future Insights : … 1) Machine Learning:

Future UAV directions emphasize improved remote sensing, crop phenotyping, infrastructure inspection, and autonomous operation through better batteries, sensors, image processing, and machine learning. Key barriers include image-processing complexity, limited energy and payload, GPS-denied navigation, and insufficient multi-UAV cooperation.

4) Future Insights :: Lightweight solar-powered batteries could extend UAV mission duration and reduce flight-planning complexity, while camera stabilization remains an unresolved remote-sensing issue.Battery weight and charging time critically affect mission duration.
4) Future Insights :: UAV remote sensing supports high-efficiency, low-cost field phenotyping in complex environments, but spectral methods show low accuracy for indirectly related complex traits.Multi-sensor systems and advanced data analysis are attracting attention for retrieving crop phenotypic traits.
D. Research Trends and Future Insights: Larger payloads, longer flight times, low-cost sensors, improved Big data image processing, and effective regulations could expand UAV-based field crop phenotyping.High-resolution UAV data is needed for accurate crop-parameter estimation under certain conditions.
A. Literature Review: UAVs enable real-time construction-site monitoring and autonomous power-line, pipeline, and tower inspection, including defect diagnosis and gas-leak detection.Applications combine onboard sensing, remote data acquisition, and infrastructure diagnostics.
B. The Deployment of UAVs for Construction & Infrastructure Inspection Applications: Industrial deployments target hard-to-reach, critical, internal, and extreme-condition assets while improving worker safety, inspection efficiency, and data processing.Examples include PG&E, AT&T, Honeywell, Maverick, and Bluestream services.
C. Challenges: Infrastructure-inspection UAVs face limited energy, short flight time, limited processing, constrained sUAV payloads, and difficulty maneuvering indoors without GPS.Payloads may include optical, TIR, color, stereo, gas-detection, and GPS sensors.
C. Challenges: Multi-UAV cooperation remains under-researched despite its potential to widen inspection scope, increase error tolerance, and accelerate task completion.The passage identifies multi-UAV cooperation as a major construction and infrastructure inspection challenge.
1) Machine Learning:: Machine learning can support autonomous UAV operation by improving conclusions, extracting features from onboard sensor measurements, and producing more concise and reliable analyses.Deep learning and CNNs are highlighted for feature extraction, image recognition, and classification.

2) Image Processing: … C. Challenges

The paper surveys UAV image-processing applications for infrastructure inspection and precision agriculture, emphasizing real-time sensing, monitoring, and automated diagnosis. It also identifies future needs and deployment constraints involving autonomy, multi-UAV cooperation, sensor limitations, weather, payload, and battery life.

2) Image Processing:: UAV image processing supports construction monitoring, infrastructure inspection, structural health monitoring, and obstacle-distance estimation using cameras, sensors, and ground-station processing.Applications include surveying construction sites, monitoring work progress, inspecting bridges and irrigation structures, detecting damage, diagnosing cracks, estimating power-line temperature, and measuring clearance from obstacles.
3) Future Insights:: Future infrastructure-inspection research should develop accurate, autonomous, real-time power-line inspection using ultrasonic sensors, TIR or color cameras, image processing, and data analysis.The goal is to monitor, detect, and diagnose power-line defects automatically.
3) Future Insights:: Multi-UAV inspections require more advanced data collection, sharing, and processing algorithms to achieve faster and more efficient operations.Future work also targets improved battery life for longer flight distance and time, alongside greater autonomy and safety in congested or indoor environments with weak GPS signals.
VII. PRECISION AGRICULTURE: UAVs support precision agriculture through crop management, weed and disease detection, irrigation scheduling, pesticide spraying, and collection of ground-sensor data.The technology is described as cost-effective and time-saving, with potential to improve crop yields, farm productivity, and profitability.
A. Literature Review: At low altitudes, UAVs provide high-precision, low-cost, real-time, high-resolution crop data for disease detection, irrigation-response analysis, weed management, and reduced herbicide use.The comparison is specifically for small crop fields relative to traditional manned aircraft.
B. The Deployment of UAV in Precision Agriculture: Precision-agriculture deployments use UAV thermal, multispectral, visual, and infrared imagery to support irrigation, disease, soil, residue, tile, maturity, and yield mapping.Examples include estimating soil moisture and crop water stress, detecting early soil-borne fungus, and predicting rice and corn yields.
B. The Deployment of UAV in Precision Agriculture: More than 95% of crop-residue-cover variability was explained by aerial thermal images, compared with 77% using visible and near IR images.UAV imagery also mapped two barley growth stages with 83.5% classification accuracy.

D. Research Trends and Future Insights … VIII. DELIVERY OF GOODS

The paper identifies machine learning, image processing, and improved UAV sensors as major research directions for precision agriculture, while also highlighting UAV delivery of goods and medical supplies. Future systems are expected to support real-time, in-field agricultural insights and monitoring.

1) Machine Learning :: Machine learning can turn UAV surveys and crop imagery into actionable crop-health insights for precision agriculture.Hummingbird uploads UAV imagery to the cloud and uses machine learning to deliver field-level recommendations.
2) Image Processing:: UAV imagery supports high-resolution farm and rangeland analysis, including crop-yield prediction, crop and weed management, and disease detection.Vegetation Indices are produced through image-processing techniques and can replace satellite or manned-aircraft imaging in some applications.
2) Image Processing:: Common vegetation indices include GVI, NDVI, GNDVI, SAVI, PVI, and EVI for agricultural image analysis.These indices have been applied to vineyards, tomato crops, and other precision-agriculture tasks.
2) Image Processing:: Rangeland vegetation-index analysis can process UAV imagery through ortho-rectification and mosaicing, plot clipping and segmentation, and hierarchical image classification.The workflow uses object-based image analysis and rule-based masking to classify homogeneous areas within 50m×50m plots.
3) Future Insights:: Relaxed flight regulations and better geo-referencing, mosaicing, and classification algorithms could expand UAV-based soil and crop monitoring.These developments are identified as future directions for precision agriculture.
3) Future Insights:: Next-generation sensors could provide onboard image processing and in-field analytics, giving farmers instant insights without cellular or cloud connectivity.The paper also calls for specialized cameras and sensors capable of real-time crop, soil, and agricultural-characteristic monitoring.
VIII. DELIVERY OF GOODS: UAVs can transport food, packages, medicines, immunizations, blood samples, and medical instruments to unreachable locations.Ambulance drones may also stream live video so paramedics can remotely observe and instruct people using medical equipment.

A. UAV-Based Goods Delivery System … A. Literature Review

The paper describes UAV delivery systems and their enabling technologies while identifying regulatory, safety, cybersecurity, weather, air-traffic, and autonomy challenges. It also surveys UAV-based real-time traffic monitoring, emphasizing detection, tracking, communication, and operational limitations.

A. UAV-Based Goods Delivery System: UAV delivery systems use GPS-equipped control processors to match package-device identifiers, transfer packages, and confirm completed deliveries.When identifiers do not match, the UAV can request the docking device identifier or network address over Bluetooth or Wi-Fi.
1) Legislation; 2) Liability Insurance: Commercial UAV delivery in the United States is constrained by licensing and line-of-sight requirements, while crashes and dropped cargo create injury, property-damage, and liability risks.UAVs can weigh up to 25 kg and approach 45 m/s.
3) Theft; 4) Weather: Delivery operations face cybersecurity threats including data theft, UAV hijacking, cargo theft, privacy invasion, and smuggling, while weather data affects safe flight planning and in-flight routing.UAV weather resistance depends on vehicle specifications, and weather data influences direction, elevation, duration, and other flight parameters.
5) Air Traffic Control; 1) Machine Learning: Large-scale delivery requires air-traffic-control coordination, predefined altitude and equipment restrictions, and autonomous systems capable of machine-learning-based route creation without GPS.Machine learning is presented as enabling UAVs to operate in unpredictable settings by recognizing encountered objects and adapting flight routes.
2) Navigation System; 3) Future Insights: GPS-independent navigation could expand delivery to remote or hazardous areas, but system-failure responses, adaptive routing, and controller handoffs remain under development.The paper identifies improved localization, low-cost sensor integration, coordination, battery endurance, collision avoidance, and aircraft-design automation as future needs.
IX. REAL-TIME MONITORING OF ROAD TRAFFIC: UAVs offer cost-effective monitoring of large or targeted road segments, but disaster-related infrastructure failures and limited battery life can disrupt control, data collection, and sampling.UAVs can complement, rather than replace, existing traffic sensors for short-duration data collection in selected areas.
A. Literature Review: Traffic-monitoring research develops UAV methods for vehicle detection, tracking, traffic-flow estimation, video relay, and real-time security-aware rerouting.Reported systems use image registration, feature extraction, shape detection, tracking, KLT, k-means, connected graphs, classifiers, simulation models, and mobile broadband.

B. Use Cases · C. Legislation

UAVs support diverse transportation and intelligent transportation system operations, including monitoring, emergency response, enforcement, roadside communication, and traffic planning. Civil UAV use is regulated through FAA registration, operating limits, authorization requirements, safety education, and penalties for violations.

B. Use Cases: Transportation applications include security surveillance, traffic monitoring, road-construction inspection, and surveys of traffic, rivers, coastlines, and pipelines.
B. Use Cases: UAVs can rapidly reach accident locations, provide advance reports, and deliver first-aid kits while rescue teams are en route.
B. Use Cases: Flying police eyes can detect traffic violations and stop vehicles by changing traffic lights or relaying messages to specific vehicles.
B. Use Cases: A UAV equipped with DSRC can act as a flying roadside unit, broadcasting warnings from locations lacking fixed roadside infrastructure.Traffic management centers can dispatch the UAV to an accident location, where it lands and warns approaching vehicles about the incident.
B. Use Cases: Other transportation uses include suspicious-behavior recognition, pedestrian-traffic monitoring, freeway monitoring, signal coordination, emergency guidance, parking analysis, and Origin-Destination flow estimation.Pedestrian data can cover demand, pedestrian characteristics, traffic flow characteristics, and walking facilities and environment.
C. Legislation: The FAA approves civil UAV use and requires small UAS weighing more than 0.55 lbs and below 55 lbs to be registered.FAA regulations are organized as prescriptive or performance-based regulations, with 20% identified as performance-based.
C. Legislation: Civil operators may obtain FAA authorization through a Section 333 Exemption or Special Airworthiness Certificate, while B4UFLY and “Know Before You Fly” provide safety information.Critical-infrastructure protections also restrict proximity to facilities such as petroleum refineries, chemical manufacturing facilities, and pipelines.
C. Legislation: FAA rules generally require visual line of sight, daylight operation, flights below 400 feet, and permission within 5 miles of an airport.Additional limits include an 87-knot ground-speed maximum, at least 3 statute miles of visibility, restrictions over unprotected people, and civil or criminal penalties for violations.

D. Challenges and Future Insights · X. SURVEILLANCE APPLICATIONS OF UAVS · A. Literature review

The paper reviews UAV surveillance applications and identifies technical, operational, regulatory, privacy, and coordination challenges that must be addressed for effective deployment. It highlights surveillance benefits such as expanded coverage and energy-efficient edge processing while emphasizing unresolved autonomy, safety, and civil-liberties concerns.

D. Challenges and Future Insights: UAVs in intelligent transportation systems face privacy, hijacking, man-in-the-middle attacks, registration, airspace regulation, and precise coordination challenges.On-board chips usually lack encryption, enabling hijacking and attacks from up to two kilometers away; registration and usage rules are also required.
D. Challenges and Future Insights: Wireless sensors, multisensor data fusion, image compression, aerial-image stitching, and network-centric infrastructure support real-time UAV control and information retrieval.These capabilities enable surveillance and live feeds for traffic control and allow operator teams to access imagery and sensor information in real time.
D. Challenges and Future Insights: Limited energy, processing, transmission range, speed, and battery endurance constrain UAV operations, motivating altitude optimization, swarming, coordination algorithms, and recharge or battery-replacement stations.Battery technology must support operational times beyond half an hour, while higher altitudes may compensate for slower speed through wider views.
D. Challenges and Future Insights: Autonomous UAV swarms must detect vehicles, UAVs, humans, and obstacles while fusing location, weather, inertial, RADAR, and LIDAR data to avoid collisions.The paper identifies swarm-intelligence algorithms and simultaneous multi-vehicle detection as major research needs.
A. Literature review: Border-surveillance studies report improved coverage, wider observation than patrols or stationary equipment, operator safety, and reduced human error, but limited detail and uncertain robustness remain.Surveillance also struggles with exact target identification and distinguishing blurred categories such as insurgents and civilians.
A. Literature review: Multi-UAV surveillance research addresses distributed task allocation, shared operational perception, cooperative perimeter monitoring, and field-tested photo mapping, although some algorithmic analyses are incomplete.Reported experiments include cooperative area surveillance, sensor deployment, fire response, load transport, and people tracking.
A. Literature review: Civil-surveillance literature finds current privacy and civil-liberties protections inadequate for complex multimodal UAV systems, recommending layered regulation alongside bottom-up privacy and ethical assessment.Inadequate mechanisms can disproportionately affect marginalized populations.
X. SURVEILLANCE APPLICATIONS OF UAVS: IoT-enabled crowd surveillance shows that MEC offloading improves energy consumption, recognition processing time, and prompt suspicious-person detection, while multiple-UAV coordination remains underanalyzed.Other practical surveillance systems integrate communication, control, sensing, image processing, and networking, including post-disaster assessment.

B. Discussion … A. Aerial Wireless Base Stations Use Cases

The discussion organizes UAV surveillance research into distinct focus areas, identifies technology trends and deployment guidance, and introduces aerial wireless base stations for resilient and expanded connectivity. Wireless-coverage use cases include ubiquitous service, gateways, relays, data collection, and satellite-integrated worldwide coverage.

State-of-the-art Research:: UAV surveillance research is classified into six categories based on focus and content, including cooperation, security, algorithms, new use cases, and products.The review summarizes advantages, disadvantages, and important concerns for these applications.
1) Research Trends:: Research trends emphasize machine learning, nano-sensors, and short-range communications to improve UAV surveillance applications.Advanced sensors support smaller size, more accurate data, and lower energy consumption, while deep learning can improve cooperation and post-processing.
1) Research Trends:: Multi-UAV cooperation can expand surveillance scope, increase error tolerance, and accelerate task completion compared with single-UAV surveillance.It requires more advanced data collection and cooperation algorithms.
2) Future Insights:: Effective UAV surveillance requires systems designed for applicable laws and budgets, customized algorithms, and field experiments across varied conditions.Recommended tests include daytime, nighttime, sunshine, and cloudy conditions to verify effectiveness and robustness.
XI. PROVIDING WIRELESS COVERAGE: UAVs can provide emergency wireless coverage as aerial base stations when cellular networks fail and can supplement ground stations for better coverage and higher data rates.The section also examines UAV links, channel characteristics, and path-loss models.
A. Aerial Wireless Base Stations Use Cases: For ubiquitous coverage, UAVs assist wireless networks with seamless service, including rapid disaster recovery and situations where cellular networks are unavailable or overloaded.These scenarios support users within the serving area.
A. Aerial Wireless Base Stations Use Cases: In remote or disaster-stricken areas, UAVs can serve as gateway nodes to backbone networks, communication infrastructure, or the Internet, and as relays between distant devices lacking reliable direct links.The gateway and relay roles extend connectivity across otherwise disconnected networks.
A. Aerial Wireless Base Stations Use Cases: UAVs can collect delay-tolerant information from distributed wireless devices, while UAV-satellite communications support integrated space-air-ground networks with high data rates and seamless wide-area coverage.Precision-agriculture sensors are an example of UAV-based data collection.

B. UAV Links and Channel Characteristics … 3) Cellular-to-UAV Path Loss Model:

The paper surveys UAV control and data links, their distinctive channel characteristics, and path-loss models for aerial communication scenarios. It highlights reliability, coverage, propagation, and modeling tradeoffs across UAV-ground, UAV-UAV, and cellular-to-UAV links.

1) Control Links:: Control links support safe UAV operation through low-latency, reliable, secure two-way communication for commands, status reports, and sense-and-avoid information.They remain important for autonomous UAV emergencies when real-time human control is unavailable.
2) Data Links:: Data links support direct mobile-UAV communication, UAV-base-station and UAV-gateway backhaul, and UAV-UAV backhaul, with capacity varying by service.Ground terminals include base stations, mobile terminals, gateway nodes, and wireless sensors.
3) UAV Channel Characteristics:: UAV communication uses UAV-ground and UAV-UAV channels, with obstacle-blocked line of sight, multipath, and high Doppler varying by channel and mobility.UAV-UAV channels are line-of-sight dominated, minimizing multipath relative to UAV-ground or ground-ground channels, while differing velocities produce high Doppler frequencies.
C. Path Loss Models: UAV path-loss models differ from terrestrial models because aerial wireless channels experience different propagation environments.Path loss represents the reduction in transmitted signal power density and is essential for wireless-channel design and analysis.
1) Air-to-Ground Path Loss for Low Altitude Platforms:: A low-altitude air-to-ground model predicts path loss statistically using more than 37000 receivers and Direct, Reflected, and Diffracted rays.The model divides receivers into three groups based on line-of-sight conditions, but assumes outdoor users represented by outdoor 2D points.
1) Air-to-Ground Path Loss for Low Altitude Platforms:: Altitude creates a path-loss tradeoff: lower altitude decreases path loss but also decreases line-of-sight probability, whereas higher altitude increases both line-of-sight probability and path loss.The model combines LOS and NLOS average losses weighted by their probabilities and includes frequency, distance, and additional-loss terms.
2) Outdoor-to-Indoor Path Loss Model:: For indoor coverage, an Outdoor-Indoor model accounts for free-space, building-penetration, and indoor losses, using ITU-certified propagation modeling.Its formulation includes UAV-to-user distance, incident angle, indoor distance, and carrier frequency, with w = 20 and g1=32.4.
3) Cellular-to-UAV Path Loss Model:: A cellular-to-UAV model estimates path loss from extensive terrestrial and aerial field samples as a function of depression angle and terrestrial coverage beneath the UAV.It includes mean terrestrial path loss, excess aerial loss, angle-dependent Gaussian shadowing, a path-loss exponent, and fitting parameters.

4) Air-to-Ground Path Loss for High Altitude Platforms: … 4) Deployment Strategies to Collect Data Using UAVs:

The surveyed work develops air-to-ground path-loss models and organizes UAV deployment strategies around transmit power, coverage, fleet size, and data collection objectives. These strategies use optimization, placement, trajectory, and clustering methods tailored to communication environments and user demands.

4) Air-to-Ground Path Loss for High Altitude Platforms:: Air-to-ground models estimate LOS probability and additional NLOS shadowing loss versus elevation angle across four built-up environments and frequencies of 2–6 GHz.The environments are suburban, urban, dense urban, and urban high-rise areas.
1) Deployment Strategies for Minimizing the Transmit Power of UAVs:: Path-loss formulations combine LOS and NLOS averages with empirical parameters, UAV–ground distance, carrier frequency, and random shadowing.The model defines P(NLOS) as 1 − P(LOS), with distance measured in kilometers and frequency in GHz.
5) UAV to UAV path Loss Model:: UAV-to-UAV channels are generally LOS-dominated, supporting free-space path loss, but continuous motion can produce high Doppler frequencies, especially for fixed-wing UAVs.The review compares path-loss models by frequency, altitude, environment, link type, experiments, and challenges.
D. UAV Deployment Strategies:: UAV deployment is formulated as a three-dimensional optimization problem whose objectives include minimizing transmit power, satisfying user rates, and selecting effective altitude and cell associations.Optimal transport theory and facility-location methods are used to derive associations and UAV locations.
2) Deployment Strategies for Maximizing the Wireless Coverage of UAVs:: Coverage-oriented placement maximizes users served or system sum-rate while accounting for antenna directionality, backhaul capacity, bandwidth, supported links, and SINR constraints.Reported approaches include coverage-region and altitude optimization, circle packing, and distributed greedy search.
3) Deployment Strategies for Minimizing the Number of UAVs Required to Perform Task:: Fleet-size strategies minimize the number of UAVs needed for area, continuous, or indoor coverage, using particle swarm optimization, energy-aware algorithms, and clustering.The indoor coverage problem is identified as NP-complete, motivating clustering-based solutions.
4) Deployment Strategies to Collect Data Using UAVs:: Data-collection deployments jointly optimize device clustering, transmit power, UAV placement or trajectories, and energy trade-offs for ground IoT and sensor networks.The surveyed methods include optimal transport, Pareto optimization, mixed-integer non-convex formulations, and successive convex optimization.

E. Interference Mitigation … 7) Future Insights:

The paper surveys interference-mitigation methods for UAV networks and identifies future research directions spanning intelligent networking technologies, wireless architectures, and unresolved deployment, mobility, energy, and security challenges.

E. Interference Mitigation: Coordinated multipoint, interference-rejection combining, network-assisted cancellation, and MIMO beamforming mitigate interference through cooperation, receiver processing, and spatial selectivity.UAVs can use more antennas than smartphones, while directional transmission or reception improves interference control.
E. Interference Mitigation: Radio-resource partitioning, optimized power control, and dedicated sky-facing cells offer complementary interference-mitigation strategies, though partitioning can waste underutilized aerial resources.Dedicated cells are especially practical in UAV hotspots with frequent and dense takeoffs and landings.
1) Cloud and Big Data; 2) Machine Learning: Cloud computing provides centralized storage, high-performance computing, big-data analysis, and network-wide monitoring, while machine learning supports autonomous settings, fault detection, intrusion detection, and behavior classification.Reinforcement learning can support decisions under unknown channel availability or base-station energy conditions.
3) Network Functions Virtualization; 4) Software Defined Networking: NFV enables programmable, seamless integration by virtualizing network functions, while centralized SDN improves radio-resource allocation, load balancing, and congestion-aware routing.Virtualizing UAVs as shared resources among cellular virtual network operators can decrease OPEX for each party.
5) Millimeter-Wave: mmWave can provide high-data-rate UAV communications, but mobility introduces rapid channel variation, blockage, directional and range constraints, and multi-user-access challenges.Intelligent cruising can move UAVs out of blockage zones and improve the probability of line-of-sight links.
6) Free Space Optical: FSO links can extend wireless connectivity to remote or sensitive areas lacking physical 3G or 4G access, supporting last-mile coverage where bandwidth and accessibility are required.UAVs can help integrate ground and aerial networks through FSO connectivity.
7) Future Insights: Future studies should address indoor UAV coverage, energy-constrained operation, experimentally validated path-loss models, absent uplink and mmWave models, and fast mmWave beam tracking.People are indoors 90% of the time, and 80% of mobile Internet access traffic occurs indoors; hovering dominates UAV energy consumption.
7) Future Insights: Open challenges include fluid network topology, routing, seamless UAV handoff, technology selection, joint D2D optimization, public-safety coordination, IoT conflicts and power control, and anti-jammer security.IEEE 802.11 is commonly used because of its wide availability and suitability for small-scale UAVs.

PART III: KEY CHALLENGES AND CONCLUSION … 1) Collision Avoidance Approaches:

The paper identifies energy management and collision avoidance as central challenges for persistent UAV missions. It surveys battery, charging, solar, machine-learning, communication, and collision-avoidance approaches used to address these challenges.

A. Charging Challenges: Battery capacity enables persistent UAV missions, but increasing capacity also increases weight and energy consumption.The literature therefore focuses on mitigating battery limitations through management and alternative energy strategies.
1) Battery Management:: Battery management research addresses mission completion through battery planning, scheduling, replacement, charge prediction, and autonomous swapping systems.Autonomous swapping systems include a landing platform, charger, storage compartment, and micro-controller.
2) Wireless Charging For UAV:: Wireless charging systems use distributed charging stations, wireless pads, solar panels, and technologies including magnetic resonance coupling and RF far-field transfer.The UAV can communicate with a control room, navigate using GPS, and proceed to an assigned station when its battery reaches a predefined level.
3) Solar Powered UAVs:: Solar-powered UAVs support long-endurance and high-altitude flights by using solar power for propulsion and batteries during nighttime or sun-absence conditions.Hybrid models combine solar, battery, and fuel-cell sources or use propulsion systems transformable between fixed-wing and quadcopter configurations.
4) Machine Learning and Communications Techniques:: Energy-aware UAV research combines communication-state management, energy-efficient networking, fleet mobility decisions, and machine learning for path planning and charging.Examples include deep reinforcement learning for fastest charging-station paths and density-based clustering for two-layer obstacle avoidance.
4) Machine Learning and Communications Techniques:: Remaining battery-recharging challenges include improving wireless charging and battery technologies, coordinating motion planning, extending research beyond multi-rotors, and applying artificial intelligence.The paper specifically highlights magnetic resonant and inductive coupling, PEM fuel cells, deep reinforcement learning, convolutional neural networks, and recurrent neural networks.
B. Collision Avoidance and Swarming Challenges: Collision avoidance is necessary because UAVs may encounter moving or stationary obstacles in indoor and outdoor environments.The survey organizes collision-avoidance research into major methodological categories.
1) Collision Avoidance Approaches:: Collision-avoidance approaches include geometric, path-planning, potential-field, and vision-based methods.Geometric methods enforce separation distances; path planning searches weighted grids for collision-free trajectories; potential fields combine attraction and repulsion; vision methods use onboard cameras and sensors.

2) Challenges: … 2) UAV New Networking Trends:

The paper identifies collision avoidance and FANET networking as major UAV challenges arising from sensing, processing, mobility, energy, and intermittent connectivity constraints. It highlights sensor fusion, onboard processing, standardization, and delay-tolerant networking as directions for safer and more reliable UAV operations.

2) Challenges:: Collision avoidance is constrained by non-cooperative obstacles, noisy or unavailable sensing data, limited onboard processing, and rapidly changing environments.ADS-B is limited to cooperative aircraft sensing, while vision-based approaches require substantial processing and work only when objects are sufficiently close.
2) Challenges:: Multi-UAV collision avoidance is complicated, with cooperative formation control and Model Predictive Controllers used to support obstacle avoidance and reduce trajectory-optimization computation time.Indoor operation adds difficulty because GPS is typically unavailable and RF signals can be reflected or degraded by obstacles and walls.
2) Challenges:: 35–70 Km/h UAV speeds require obstacle avoidance to execute quickly, while limited power and payload constrain sensor weight, size, and power requirements.Typical sensors include IR, ultrasonic, laser scanner, LADAR, and RADAR, which can be heavy and large for sUAVs.
3) Future Insights:: Future work emphasizes multisensor collision avoidance, autonomous hovering, smoother paths, optimized energy consumption, global standardization, and low-power onboard processing.Recommended sensors include vision, laser range finder, IR, and ultrasound; powerful low-power processors are needed for dynamic sense-and-avoid, path replanning, and image processing.
C. Networking Challenges: FANETs require new communication protocols because 3D mobility, speeds of 30 to 460 km/h, fluid topology, and rapidly changing link quality challenge multi-UAV communications.Wireless communication is needed among UAVs, ground control stations, and satellites despite frequent connection interruptions and limited energy resources.
1) FANET Challenges: •: FANET power constraints limit computation, communication, and endurance, while complex hardware configuration makes network resource management difficult; energy-aware deployment, LLT, SDN, and NFV are proposed responses.These approaches address energy limitations and configuration difficulties in the FANET environment.
2) UAV New Networking Trends:: Ad-hoc FANETs face high mobility, long distances, fluid topology, link delays, and high channel error rates, causing transmitted data to be lost or delayed; DTN architecture addresses these disruptions.DTN uses store-carry-forward routing to compensate for intermittent connectivity and long packet-delivery delays.
2) UAV New Networking Trends:: DTN can support reliable data transport through Bundle Protocol, Licklider Transmission Protocol, and CFDP, including hybrid routing that combines aerial DTN with ground AODV.UAVs store, carry, and forward messages until connectivity or acknowledgment enables delivery.

b) Network Function Virtualization (VFV): … 1) Attack Vectors in UAV Systems:

The section surveys virtualized, software-defined, and low-power networking approaches for FANETs, identifies future networking priorities, and categorizes UAV-system cyberattacks across communications, vehicles, ground stations, and humans.

b) Network Function Virtualization (VFV):: NFV virtualizes network devices, hardware, and functions through IT virtualization technologies, enabling programmable FANET networking and replacing physical components with virtual appliances.NFV changes how network functions are provisioned and can reduce network management burden.
b) Network Function Virtualization (VFV):: NFV- and SDN-based UAV applications include multi-tier drone-cell networks complementing terrestrial cellular systems and swarm-based video monitoring platforms in rural FANETs.The drone-cell management framework uses UAVs as aerial base stations, while VMPaaS uses UAV swarms.
c) Software-Defined Networking (SDN):: SDN separates control and data planes and provides centralized programmability with a global network view, simplifying deployment, management, and reconfiguration of applications and services.SDN can be consolidated with NFV to address NFV’s interconnection and control complexity.
c) Software-Defined Networking (SDN):: SDN helps FANETs manage rapid topology changes, intermittent links, UAV outages, coverage limitations, battery recharging, and overall network-management complexity.OpenFlow implementations may use centralized, decentralized, or hybrid control planes, but controller connectivity and current topology knowledge must be maintained.
d) Low power and Lossy Networks (LLT):: FANETs face limited battery, memory, and computational resources, motivating low-power and lossy-network approaches for handling UAV power constraints.RPL, low-power WiFi, IEEE 802.15.4, and 6LoWPAN support constrained networking and UAV-UAV communications.
3) Future Insights:: Future work should develop SDN services for FANET surveillance, safety, and security, while improving deployment reliability, reachability, and support for fast UAV mobility.These priorities are explicitly identified as future insights for FANET networking.
3) Future Insights:: Future networking research should address SDN security protocols, fully decoupled switching devices, NFV virtual topologies, customized end-to-end protocols, dynamic embedding, energy conservation, bandwidth, and quality of service.The section also calls for FANET-specific routing protocols and new networking models because existing MANET protocols may provide unreliable UAV communications.
D. Security Challenges: UAV systems have a large attack surface spanning communications links, UAVs, ground control stations, and human operators, creating substantial cybersecurity challenges.Attacks include eavesdropping, session hijacking, spoofing, identity hacking, malware, viruses, key loggers, and social engineering.

2) Taxonomy of Cyber Security Attacks/Challenges Against UAV Systems: · 3) Literature Review of the State-of-the-art Security Attacks/Challenges and Countermeasures:

The paper classifies UAV cyber-security challenges into confidentiality, integrity, and availability, then organizes related research into specific attack studies, general security analyses, and security-framework development. Reviewed countermeasures address attacks including GPS spoofing, DDoS, sensor-input spoofing, hijacking, authentication, and key management, but several approaches retain scope and deployment limitations.

2) Taxonomy of Cyber Security Attacks/Challenges Against UAV Systems:: UAV cyber-security attacks are organized around confidentiality, integrity, and availability challenges.Confidentiality protects information from unauthorized access, integrity protects authenticity, and availability ensures services and relevant data remain accessible to authorized users.
2) Taxonomy of Cyber Security Attacks/Challenges Against UAV Systems:: Confidentiality attacks include malware, hijacking, and social engineering, while integrity attacks can modify or fabricate collected data and issued commands, including through GPS signal spoofing.Availability attacks include DoS methods such as flooding communications links, overloading processing units, or depleting batteries.
3) Literature Review of the State-of-the-art Security Attacks/Challenges and Countermeasures:: The literature review distinguishes specific attack discussions, general security analyses, and security-framework development.Framework work introduces monitoring systems, simulation test beds, or anomaly-detection frameworks, and Table XIX summarizes 15 reviewed papers.
3) Literature Review of the State-of-the-art Security Attacks/Challenges and Countermeasures:: GPS spoofing research analyzes UAV hijacking vulnerabilities, exploitation, countermeasures, risk mitigation, and impacts, including cryptography-based signal authentication and multiple receivers.The cited anatomical investigation aims to provide practical insights for UAV-system practitioners.
3) Literature Review of the State-of-the-art Security Attacks/Challenges and Countermeasures:: Specific-attack studies also address DDoS, sensor-input spoofing, and hijacking through attack modeling, mitigation, and countermeasures for MitM and command-injection vulnerabilities.Sensor-input spoofing can create a new control channel when an attacker understands the sensor algorithms and manipulates the victim’s environment.
3) Literature Review of the State-of-the-art Security Attacks/Challenges and Countermeasures:: A flight-pattern fingerprinting scheme detects all direct hijacking scenarios by comparing simulated scenarios with a baseline flight profile.The approach can experience temporary control instability, including short-term amplitude decreases.
3) Literature Review of the State-of-the-art Security Attacks/Challenges and Countermeasures:: Authentication and key-management research modifies authentication, key-agreement, and handover protocols for LTE-integrated CNPC networks.A separate secure-link scheme is constrained because its symmetric key is generated by a trusted third-party computer and hard-coded in source code, limiting feasibility.
3) Literature Review of the State-of-the-art Security Attacks/Challenges and Countermeasures:: Security frameworks include UAVSim for configurable cyber-security experiments and detection systems using in-flight behavior, Mahalanobis distance, and Message Dropping Rate.Reported limitations include UAVSim’s initial attack library containing only jamming and DoS attacks, non-detectable attacks that do not alter in-flight data, and single-node rather than cooperative detection.

4) Summarizing of Cyber Security Challenges for UAV Systems: · XIII. CONCLUSION · XIV. ACRONYMS

The survey identifies dominant UAV cybersecurity threats, limitations in current defenses and simulation environments, and broad civil-application opportunities requiring further research and regulation. It concludes with an acronym glossary supporting the survey’s terminology.

4) Summarizing of Cyber Security Challenges for UAV Systems:: DoS and hijacking attacks involving signal spoofing are the most prevailing UAV-system threats, causing serious availability issues and severely damaging system behavior.Suggested countermeasures include strong authentication, signal-distortion detection, and direction-of-arrival sensing.
4) Summarizing of Cyber Security Challenges for UAV Systems:: Existing countermeasure algorithms mainly address single UAV systems, leaving multiple-UAV scenarios without concrete cooperative security solutions.A proposed direction is adapting distributed security frameworks such as Kerberos for multiple UAV systems.
4) Summarizing of Cyber Security Challenges for UAV Systems:: Security analyses often overlook hardware and software differences, while countermeasures developed for other communication systems may face deployment difficulties across UAV platforms.Some attacks exist only under specific hardware or software configurations.
4) Summarizing of Cyber Security Challenges for UAV Systems:: Current UAV security simulation test beds remain immature because emulators cover few attack scenarios and specific hardware/software configurations.Possible improvements include powerful simulation tools such as Labview or customized simulation environments.
XIII. CONCLUSION: UAVs are becoming ubiquitous across civil applications, supporting dangerous or inaccessible tasks through real-time monitoring, rescue, delivery, sensing, and infrastructure inspection.The survey reviews these applications, their challenges, research trends, and future uses.
XIII. CONCLUSION: $45 Billion is the forecast addressable UAV market value, with civil infrastructure expected to dominate through construction monitoring and power-line and gas-pipeline inspection.These applications can reduce work injuries, inspection costs, and time compared with conventional methods.
XIII. CONCLUSION: Further research is needed for practical UAV delivery, higher-altitude traffic monitoring, efficient multi-UAV surveillance and networking, and solutions to charging, swarming, collision avoidance, and security challenges.The survey also calls for a complete legal framework and institutions regulating civil UAV use globally.
XIV. ACRONYMS: The acronym section lists abbreviations and definitions used throughout the survey, including SAR for Search and Rescue, SDN for Software-Defined Networking, and SHM for Structural Health Monitoring.The glossary also includes terms such as RSU, SAVI, and SOC.