Research of UAV and sensor network integration features for routing optimization and energy consumption reduction⋆ Nadiia Dovzhenko1,†, Yevhen Ivanichenko1,†, Pavlo Skladannyi1,2,*,† and Oleksii Zhyltsov1,† 1 Borys Grinchenko Kyiv Metropolitan University, 18/2 Bulvarno-Kudryavska str., 04053 Kyiv, Ukraine 2 Institute of Mathematical Machines and Systems Problems of the National Academy of Sciences of Ukraine, 42 Ac. Glushkov ave., 03680 Kyiv, Ukraine Abstract Modern unmanned aerial vehicles (UAVs) are increasingly integrating with sensor networks, significantly expanding the capabilities of real-time data collection, transmission, and processing. This integration is critically important for various sectors, including environmental monitoring, smart city infrastructure management, agriculture, and military operations. UAVs provide mobility and access to remote and hard- to-reach locations, enabling effective monitoring in areas where traditional networks are unavailable or ineffective. However, alongside these advantages, numerous technical challenges arise. These challenges include optimizing UAV flight routes to ensure maximum sensor network coverage, minimizing energy consumption, and addressing data security issues such as cyber threats. Another important aspect is flight duration, which depends on UAV battery capacity and energy-saving methods for sensor nodes, especially through the use of alternative energy sources, such as solar panels. The study presents a model of dynamic interaction between UAVs and a sensor network, examining the process of data collection, and transmission to a central server, and the impact of increasing the number of sensor nodes on the overall mission time. A stochastic model is proposed to account for environmental heterogeneities, such as data transmission delays caused by obstacles or changes in connection speed. An analysis is conducted to evaluate the impact of these factors on data collection efficiency and to optimize flight routes, with a focus on dynamic programming algorithms and heuristic methods. Keywords UAVs, drones, sensor networks, IoT, nodes, energy efficiency, routing, security, reliability, connectivity, data 1 1. Introduction UAVs can serve as platforms for meteorological measurement systems. Unmanned aerial vehicles (UAVs) are increasingly being They have advantages over fixed-wing UAVs, whose used in various fields of human activity. For example, in high speed limits spatial and temporal resolution, making agriculture, GPS-guided UAVs are employed for spraying them less sensitive to turbulent processes [3]. crops in the fields. The use of UAVs significantly saves Today, relatively affordable multicopters are available, resources (such as time and chemicals) and ensures accurate capable of lifting payloads of 3–5 kg to altitudes of 2–4 km and precise treatment of agricultural lands compared to with flight durations of 30–40 minutes. Modern UAVs, manned aviation [1]. equipped with onboard navigation and control systems, In some European Union countries, drones are even perform a wide range of functions. For example, UAVs can used for customer deliveries. During military conflicts and be programmed with fixed flight routes (using coordinates, wars, UAVs are utilized for delivering medications, altitude values, and specific waypoints) and can change humanitarian aid, and combat supplies to hard-to-reach their routes or return to the starting point upon command areas. Certain drone models are also employed to inspect from the ground control station. Additionally, UAVs can fly power lines, transformers, and pipelines [2]. over designated points, collect and transmit telemetry Emergency services deploy drones for monitoring, information about flight parameters and the operation of forecasting, and controlling hazardous sites, contributing to target equipment, and provide software control for this both safety and environmental protection. In particular, equipment [4]. CPITS-II 2024: Workshop on Cybersecurity Providing in Information 0000-0003-4164-0066 (N. Dovzhenko); and Telecommunication Systems II, October 26, 2024, Kyiv, Ukraine 0000-0002-6408-443X (Y. Ivanichenko); ∗ Corresponding author. 0000-0002-7775-6039 (P. Skladannyi); † These authors contributed equally. 0000-0002-7253-5990 (O. Zhyltsov) nadezhdadovzhenko@gmail.com (N. Dovzhenko); © 2024 Copyright for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0). y.ivanichenko@kubg.edu.ua (Y. Ivanichenko); p.skladannyi@kubg.edu.ua (P. Skladannyi); o.zhyltsov@kubg.edu.ua (O. Zhyltsov) CEUR Workshop ceur-ws.org ISSN 1613-0073 236 Proceedings Moreover, modern UAVs are actively integrated into fields “flying wings”. These aircraft usually have high- such as environmental monitoring, terrain mapping, and mounted wings, often in a V-shape, with electric construction. Drones are also used in search and rescue motors. They may also feature more complex operations and to assess agricultural land conditions. fuselage designs—from gondolas to single-fuselage Advancements in artificial intelligence, sensor networks, solutions. They are equipped with piston engines and microchips have significantly improved the autonomy and typically take off from specially designed and accuracy of UAV flights, making them indispensable launch platforms. Their landing is accomplished across many industries. This enables greater mobility and either by parachute or via traditional aircraft efficiency in operations while maintaining relatively low methods. The advantage of such drones is their operational costs for UAVs [5]. ability to perform much longer and more complex missions, remaining airborne for up to 5 hours. 2. Evolution, classification, and  Medium UAVs have heavier landing gear and modern applications of unmanned more complex takeoff and landing systems, which are more similar to traditional aviation principles. aerial vehicles These drones can perform long-duration flights Research in the field of unmanned aerial vehicles (UAVs) (approximately up to 20 hours) and can ascend to has a long and rich history. It began during World War I, in altitudes of up to 6 kilometers. 1917–1918, with developments in the United States and the  Heavy UAVs are designed as “air giants” compared United Kingdom. One of the prototypes, the Kettering Bug, to other UAVs. They can reach altitudes of up to was an early cruise missile with a simple control system that 20 kilometers and remain airborne for more than allowed it to fly along a predetermined route. Another 24 hours. The construction of such UAVs involves example, is the Aerial Target, an unmanned aircraft the use of sufficiently complex and powerful controlled via radio, which was developed for anti-aircraft engines and landing gear, making them effective artillery training. Although the project faced numerous for strategic missions and multifunctional tasks. technical limitations and was ultimately unsuccessful, it laid the foundation for further advancements in aviation. It is also appropriate to classify UAVs by weight, as this During World War II, the German army deployed the represents a separate classification for determining the first attack unmanned aerial vehicle (UAV)—the V-1 flying capabilities of such devices. Small UAVs (weighing up to 5 bomb. Later, researchers classified this aircraft, and its kg) can perform short-term reconnaissance missions, such predecessors, as cruise missiles rather than conventional as target detection in hard-to-reach areas. Research UAVs. However, it is important to note that their indicates the potential for the weight of heavy UAVs to characteristics laid the groundwork for modern UAVs, increase to 15 tons in the coming decades [7]. particularly in terms of autonomy and the ability to These options will be capable of performing strategic accurately reach targets without a pilot. functions, combining reconnaissance, monitoring, and the From the 1950s to the 1970s, significant scientific management of a large amount of equipment for various research and development were conducted in the field of tasks [8]. combat UAVs, particularly in versions capable of flying at high altitudes, remaining airborne for extended periods, and 3. Technical aspects of integrating conducting surveillance. UAVs such as the Ryan Firebee, sensor networks and UAVs AQM-34 Ryan Model 147, and Teledyne Ryan AQM-91 Firefly were not only used for reconnaissance and training The world of the Internet of Things (IoT) is broad and tasks but also represented advancements in remote control multifaceted, encompassing a wide range of industries, each technologies that eventually became fully operational with its own unique features and technological combat platforms. requirements. However, it is more appropriate to view IoT Modern drones are used not only for military purposes, not as a single technological domain, but as a combination reconnaissance, and precision strikes but also for civilian of different concepts, protocols, and technologies that vary tasks, such as search and rescue operations, infrastructure depending on the application. monitoring, agriculture, environmental monitoring, and Sensor networks are a key component of IoT, providing cybersecurity. UAVs such as the RQ-1 Predator, MQ-9 monitoring of physical parameters of the environment. Due Reaper, DJI Phantom, MQ-25 Stingray, Bayraktar TB2, to limited resources and the need to operate in harsh Hermes 900, Wing Loong II, XQ-58A Valkyrie, and others conditions, these networks often face challenges such as have significantly expanded their capabilities thanks to the node failures and malfunctions, leading to new issues, continuous development of artificial intelligence particularly in scaling the network to accommodate large technologies, improvements in computing power, and the numbers of connected devices, processing large volumes of implementation of sensor systems [6]. data, and ensuring security. Analyzing the existing varieties of unmanned aerial Therefore, a logical step in the development of these vehicles, they can generally be classified by their structural technologies is the integration of UAVs with sensor features. For example: networks, opening up new possibilities for efficient data collection, monitoring, and management in various areas of  Small UAVs are typically built based on classic human activity, including the deployment of “smart” cities, aerodynamic designs, with variations such as environmental monitoring, critical infrastructure systems, cybersecurity, and even military applications. 237 One of the main advantages of using UAVs as mobile base networks. The use of modern encryption and authentication stations is the improvement of communication between mechanisms helps prevent unauthorized access to data, sensor nodes and drones, reducing signal loss, increasing which is especially important for military and industrial the likelihood of direct line of sight, and, most importantly, applications [12]. decreasing the energy consumption of sensor resources. It is worth noting that sensor networks consist of hundreds or 4. Model of dynamic interaction thousands of nodes that collect, process, and transmit data to central servers or cloud platforms using wireless between UAVs and sensor communication protocols for further analysis. networks Therefore, reducing energy consumption is especially The integration of UAVs and sensor networks offers vast important for nodes with low battery power, as it helps opportunities to improve efficiency and security in many extend their operational life [9]. sectors, but it also requires addressing several technical To ensure effective data collection and transmission challenges to ensure the stable and reliable operation of such between UAVs and sensor nodes, technologies such as systems. LoRaWAN, Zigbee, and 5G are used. A model of the dynamics of data collection and For comparison, LoRaWAN and Zigbee provide stable interaction between UAVs and a sensor network is proposed, communication in difficult conditions with low power based on a specific scenario. In this scenario, the drone flies consumption, extending the operational life of the sensors. over several sensor nodes, collecting data from them and 5G technology enables the transmission of large transmitting it to a central server. For modeling this process, amounts of data in real time and ensures low latency, which a sensor network consisting of 50, 100, 200, or 500 nodes is is critical for tasks that require an immediate response. considered. The time for data collection, data transmission, Additionally, drones can be used to install sensor nodes and command processing is also taken into account. in hard-to-reach or dangerous locations for humans, such as For example, the flight time to the sensor network nodes disaster zones, mountainous regions, seismically active is determined by formula (1) and depends on the distance areas, or sites of industrial accidents. In this case, the use of between the sensor nodes and their quantity. UAVs as mobile platforms for sensor networks allows for 𝑡 = , (1) flexible positioning, data collection, and preliminary processing over large areas, providing continuous 𝑑 is the distance between sensor nodes (e.g., 100 m), 𝜐 is monitoring and rapid response to real-time changes [10]. the speed of the drone (for example, 10 m/s). If the number Another key aspect of integrating sensor networks and of nodes increases but the area size remains unchanged, the UAVs is the optimization of drone flight paths to ensure distance between the nodes decreases proportionally. maximum sensor network coverage while minimizing This, in turn, will affect the flight time, reducing it, but energy consumption. Key parameters that affect the increasing the number of nodes the drone interacts with and efficiency of data collection include flight speed, distance to connects to [13]. sensor nodes, transmitter power, flight altitude, and more. The total interaction time between a sensor node and The use of dynamic programming, heuristic algorithms, the drone is calculated as follows: and machine learning methods enables the real-time 𝑡 =𝑡 +𝑡 +𝑡 , (2) optimization of drone routes, improving the performance of where 𝑡 is time for data collection from a the data collection system and reducing the likelihood of single node, 𝑡 is time for data transfer from a errors or failures. single node, and 𝑡 is flight time. However, one of the biggest challenges remains the The data transfer rate is determined as follows: limited capacity of drone batteries, which determines their 𝑡 = , (3) capabilities and flight duration. Due to several factors, lithium-ion batteries remain the most efficient; however, where 𝑅 is data volume from one node (for example, 10 the issue of limited energy forces the search for new MB), and 𝑆 is transfer speed (for example, 1 Mbit/s). approaches and solutions. For example, new types of To formulate the mathematical model of the interaction batteries, fuel cells, or hybrid power sources may be between sensor nodes and UAVs (drones) in such a scenario, considered for drones. For sensor nodes, solar panels are a system of equations can be used to describe the process of used, reducing the frequency of recharging and ensuring data collection, transmission, and processing from sensors, continuous system operation. Energy-saving methods for taking into account discrete time intervals. sensor components, such as adaptive module shutdowns The model will be based on stochastic process concepts and optimization of data collection frequency, are also to simulate random delays and heterogeneities [14]. actively being researched [11]. The model for drone data collection and transmission Additionally, to ensure proper synchronization between can be described as a discrete process that defines the sensor nodes and drones, especially in complex or dynamic change in system state at the time 𝑡 , when the drone moves environments, special routing algorithms, and dynamic data between sensors and collects information: correction are used. Reliable synchronization is critically 𝑋, = 𝑋 , + 𝜇 , 𝐿 , (𝑋 , 𝜉 , ), (4) important to prevent data loss and ensure the stable where 𝑋 , is the volume of collected data at a given time 𝑡 , operation of the entire system. 𝜇 , is discretization step parameter, which regulates data It is also worth noting the importance of cybersecurity, state changes, 𝐿 , (𝑋 , 𝜉 , ) is a function that describes the which is another key aspect of integrating UAVs and sensor process of data collection and transmission from sensor 238 node 𝑖 under certain conditions, 𝜉 , is models random delays or other changes in the system. For a system with 𝑁 sensor nodes, the total data collection time will be defined as: 𝑇 =𝑁∗𝑡 , (5) where 𝑁 is the number of sensors, and 𝑡 is the total interaction time with one sensor node. The overall interaction time of the drone with the sensor network increases almost linearly with the number of sensor nodes, as shown in Fig. 1. Each additional sensor node increases the total mission time due to the time required for flight, data collection, and data transmission to the gateway or server. With a significant increase in the number of sensor nodes, optimization strategies should be considered, such as Figure 2: Dependence of total data collection time on the using multiple drones in parallel, dividing areas of number of sensor nodes, considering heterogeneities in responsibility, or using faster data transmission channels. flight, data collection, and transmission times To optimize the route between sensor nodes, graph- based approaches or the traveling salesman problem can be Fig. 2 shows the heterogeneities accounted for in the time used, where the drone must find the most efficient route for flight, data collection, and transmission. that minimizes the distance between nodes. It can be argued that adding random variations in time simulates real conditions, where delays due to obstacles or changes in data transmission speed may occur [15]. 5. Data Security Challenges in UAV- Integrated Sensor Networks Since sensor networks and their components are often deployed in uncontrolled or insufficiently protected physical environments, especially in the case of integration with UAVs, it is crucial to pay particular attention to the impact of attacks and threats on both individual nodes/sensors and drones. Due to the numerous advantages of UAVs interacting with sensor nodes, there is an increased risk of unauthorized access, data manipulation, or modification, as well as heightened vulnerability to the compromise of nodes and drones through physical access to network elements. In Figure 1: Dependence of total data collection time on the such cases, malicious software can be implanted, potentially number of sensor nodes compromising the integrity and confidentiality of the processed information. The aforementioned threats From Fig. 1, it can be observed that as the number of sensors jeopardize not only individual network components but also increases, the total time increases linearly, as the drone the security of the entire system [16]. needs more time to fly over all the sensors, collect data, and Today, there is a broad range of threats and types of transmit it to the server. attacks that can target sensor networks, occurring at various levels. One of the most common types of attacks is However, the presented calculations do not always jamming, aimed at introducing additional noise and reflect realistic scenarios. In real-world conditions, interference in the physical channel for wireless signal heterogeneities may arise, such as delays due to obstacles transmission, which can disrupt the correct interaction (e.g., trees, buildings, or other objects that may slow down between nodes and UAVs. Other common threats include the drone or cause additional energy expenditure), changes physical interference with network operations, sensor in data transmission speed (sometimes the speed may spoofing, or attacks on information leakage through direct fluctuate due to interference, the distance between drones access to network components. Such threats may lead to and the server, etc.), or variations in flight time due to unpredictable consequences, including modification, delay, differing distances between sensors (sensors may be or loss of data sent to gateways, routers, or central servers, distributed unevenly), and so on. potentially causing misinformation or improper data processing. To account for such heterogeneities, random delays or For instance, at the link layer, a significant threat variables can be added to the flight time, data collection, and includes collision attacks, where identical frequency information transmission time, for example: channels are used. Such attacks lead to resource depletion 𝑑 of nodes by forcing them to repeatedly retransmit damaged 𝑡 = +𝜖 , (6) 𝜐 or lost packets to recipient nodes, ultimately negatively where 𝜖 is a random value of delay or change that creates affecting network performance. When interacting with heterogeneity. UAVs, these attacks can trigger excessive energy consumption by nodes, reducing the overall system uptime 239 and increasing the risk of node disconnections from the relevance, network self-organization capabilities, time network. synchronization, as well as node tracking and security At the network level, attackers can alter or spoof incident localization. In the case of UAV integration, these routing data, redirecting legitimate traffic to compromised aspects become critically important, as the constant nodes. An example is the Black Hole attack, in which a node movement of drones imposes additional requirements on intercepts data and does not forward it, creating a "black the system’s response time to potential threats. hole" in routing. Another example is the Selective Therefore, ensuring the security of wireless sensor Forwarding attack, where a malicious node selectively networks when used with UAVs demands a comprehensive forwards only part of the packets, ignoring the rest. During approach that includes multi-level protection against interactions with UAVs, these attacks can cause serious various types of attacks, adaptive resource management, disruptions in data delivery and negatively affect the and continuous improvement of security mechanisms to timeliness of data receipt. effectively counter emerging cyber threats [17]. Additionally, sensor networks are vulnerable to eavesdropping and traffic analysis attacks, which allow 6. Conclusions attackers to access confidential information, modify it, or alter it for further attacks on the network. To mitigate risks The integration of unmanned aerial vehicles (UAVs) with associated with these threats, robust encryption and sensor networks opens new opportunities for efficient data authentication methods must be employed to protect data collection and transmission across various industries, transmitted between UAVs and sensor nodes. particularly in remote and hard-to-reach locations. The Thus, ensuring the security of sensor network results presented in this study indicate that such integration components, especially when used in conjunction with significantly enhances the quality of real-time monitoring unmanned aerial vehicles (UAVs), is a highly challenging and process management. task due to the limited resources of each sensor. The The proposed mathematical model describes a linear constrained processing power, memory, and energy increase in total mission time as the number of sensor nodes resources of sensor nodes create challenges for effective grows, underscoring the need for optimizing UAV flight encryption key management, which is essential for data routes. Such optimization reduces interaction time with protection. sensor nodes and improves data collection efficiency. Modern strategies require the development of The study also examines the implementation of distribution and key-update mechanisms that can adapt to alternative power sources, such as solar panels for sensor rapid network topology changes while maintaining a high nodes and hybrid batteries for UAVs. These solutions level of security amidst constant UAV interaction. It is positively impact the system's continuous operation time worth emphasizing that the dynamic nature of this topology and extend flight durations. also places additional demands on the speed of adaptation Additionally, special attention is given to data security, of security mechanisms. as the use of UAVs increases the risk of unauthorized access, As the number of sensor nodes increases, the load on manipulation, and attacks on the sensor network. data storage and transmission systems also rises, Implementing multi-level protection mechanisms, adaptive complicating security maintenance. The use of data resource management, and advanced encryption methods compression technologies and selective data transmission ensures data protection and system stability amid algorithms can optimize network performance, reducing the continuous UAV interactions. volume of transmitted data and thereby shortening the period during which the network remains vulnerable to attacks. References However, resource limitations necessitate additional [1] Z. Li, et al., An Adaptive and Automatic Power Supply approaches to achieve comprehensive protection for sensor networks and UAVs. Distribution System with Active Landmarks for To ensure protection at every level of the system, it is Autonomous Mobile Robots, Sensors 24 (2024). doi: crucial to enable interaction between security mechanisms. 10.3390/s24186152. For instance, effective energy resource management at the [2] J. Deng, et al., A Methodology to Monitor Urban channel level can significantly reduce vulnerability to Expansion and Green Space Change Using a Time resource-depleting attacks, such as Denial of Service (DoS) Series of Multi-Sensor SPOT and Sentinel-2A Images, attacks. For this purpose, it is beneficial to implement Remote Sens. 11 (2019) 1230. doi: 10.3390/rs11101230. strengthened authentication and encryption protocols that [3] N. Dovzhenko, et al., Comprehensive Analysis of provide additional protection against unauthorized access at Efficiency and Security Challenges in Sensor Network the physical and network levels. Routing, in: Cybersecurity Providing in Information Modern encryption methods and advanced security algorithms at the network level allow for the prevention of and Telecommunication Systems II, vol. 3550 (2023) routing data spoofing and protect the system from Black 275–280. Hole and Selective Forwarding attacks. At the transport [4] N. Cheng, et al., AI for UAV-Assisted IoT level, Flooding attacks and similar strategies targeting the Applications: A Comprehensive Review, IEEE depletion of node memory and computational resources Internet of Things Journal (2023). doi: represent another threat that must be taken into account. 10.1109/JIOT.2023.3268316. Additionally, the threat of synchronization attacks [5] K. P. Valavanis, G. J. Vachtsevanos, Handbook of should be considered, as such intrusions can disrupt data Unmanned Aerial Vehicles, Springer Publishing transmission and interfere with coordinated operations Company, Incorporated (2014). between sensors and UAVs. [6] J. M. Maddalon, et al., Perspectives on Unmanned In addition to the core information security goals (confidentiality, integrity, and availability), sensor networks Aircraft Classification for Civil Airworthiness also require secondary security objectives, such as data Standards (No. NF1676L-16131) (2013). 240 [7] M. Asadpour, et al., Micro Aerial Vehicle networks: An Experimental Analysis of Challenges and Opportunities, IEEE Communications Magazine 52(7) (2014) 141–149. doi: 10.1109/MCOM.2014.6852096. [8] N. Dovzhenko, et al., Method of Sensor Network Functioning under the Redistribution Condition of Requests between Nodes, in: Cybersecurity Providing in Information and Telecommunication Systems, vol. 3421 (2023) 278–283. [9] M. Mozaffari, et al., A Tutorial on UAVs for Wireless Networks: Applications, Challenges, and Open Problems, IEEE Communications Surveys & Tutorials, 21(3) (2018) 2334–2360. doi: 10.1109/COMST.2019. 2902862. [10] E. Y. Song, G. J. FitzPatrick, K. B. Lee, Smart Sensors and Standard-based Interoperability in Smart Grids, IEEE Sensors J. 17(23) (2017). doi: 10.1109/JSEN.2017.2729893. [11] A. Hassanien, et al., Signaling Strategies for Dual- function radar Communications: An Overview, IEEE Aerospace and Electronic Systems Magazine, 31(10) (2016) 36–45. doi: 10.1109/MAES.2016.150225. [12] R. W. Beard, T. W. McLain, Small Unmanned Aircraft: Theory and Practice, Princeton, NJ, USA: Princeton Univ. Press (2012). doi: 10.1515/9781400840601. [13] G. Liu, et al., Joint radar communication system design based on filter bank multicarrier modulation scheme, IET Radar, Sonar & Navigation, 17(1) (2022) doi: 10.1049/rsn2.12323. [14] R. Beard, et al., Autonomous Vehicle Technologies for Small Fixed-Wing UAVs, J. Aerospace Comput. Inf. Commun. 2(1) (2005) 92–108. doi: 10.2514/1.8371. [15] O. Barabash, et al., Development of a hybrid network traffic load management mechanism using smart components, in: IEEE 7th International Conference on Methods and Systems of Navigation and Motion Control (MSNMC) (2023) 38–41. [16] Z. Hu, et al., Analytical Assessment of Security Level of Distributed and Scalable Computer Systems. International Journal of Intelligent Systems and Applications, vol. 8, no. 12 (2016) 57–64. [17] P. Openko, et al., Zabezpechennia nadiinosti ta bezpeky u suchasnykh bezprovodovykh sensornykh merezhakh na osnovi vprovadzhennia metryky RSSI. Povitriana mits Ukrainy, 1(6) (2024) 131–136. doi: 10.33099/2786-7714-2024-1-6-131-136 241