Adaptive dynamic programming–based fault-tolerant control of multi-UAV formation system

The Internet of Things (IoT) can be supported by multiple isolated sensor networks (MISN) when the connectivity of IoT is constrained by obstacles. Mobile nodes have been used as relays to connect partitioned networks. However, mobile nodes have the problem of not being able to move or moving slowly when encountering obstacles. In order to tackle this issue, three types of Unmanned Aerial Vehicle (UAV) swarm modes are proposed, and UAV swarm assisted connectivity enhancement algorithms (UsCE) are designed. A UAV swarm with high degree of freedom and flexibility provides a new way in IoT for solving the above connectivity problem. Our target is to find an optimal solution that minimizes the number of UAVs in the swarm and maximizes the connection time of MISN. We divide the working modes of a UAV swarm into hovering and flying. Firstly, the ground sink nodes are classified by a MISN's sink node classification algorithm to generate hovering points for the UAV swarm. Secondly, the results are optimized and adjusted by a minimum UAV swarm hovering connection algorithm to obtain an optimal solution under the hovering mode. Finally, we achieve an optimal connectivity when the UAV swarm works in flying mode through the UAV swarm flight connectivity algorithm and compare it with two previously proposed algorithms. Simulation results show that the complexity of the algorithms is low, the connection time of MISN increases significantly, and the number of UAVs is small. An optimal UAV swarm assisted connectivity enhancement scheme for MISN of different scales is derived.

For the large-scale operations of unmanned aerial vehicle (UAV) swarm and the large number of UAVs, this paper proposes a two-layer task and resource assignment algorithm based on feature weight clustering. According to the numbers and types of task resources of each UAV and the distances between different UAVs, the UAV swarm is divided into multiple UAV clusters, and the large-scale allocation problem is transformed into several related small-scale problems. A two-layer task assignment algorithm based on the consensus-based bundle algorithm (CBBA) is proposed, and this algorithm uses different consensus rules between clusters and within clusters, which ensures that the UAV swarm gets a conflict-free task assignment solution in real time. The simulation results show that the algorithm can assign tasks effectively and efficiently when the number of UAVs and targets is large.

Unmanned aerial vehicle (UAV) swarms offer a resource-effective and time-efficient method of collecting and analyzing data for a range of applications. The paper presents a cutting-edge collective artificial intelligence-powered self-organizing UAV swarm mission planning and preparing platform designed to enhance task performance with a fleet of UAVs and facilitate terrain monitoring. Users can create requests easily from the QT interface due to the cloud-based multi-user platform's interactive capabilities, which enable smooth user collaboration and real-time video viewing for group study of dynamic landscape imagery. Optimizing the behavior and performance of UAV swarm navigation maps, the UAV map configurator makes it easier to create and modify them. Additionally, the QT service layer guarantees safe data transfer to cloud servers, and the parameter gossip system encourages coordination and communication among swarm members. This integrated data sets significant parameters such as the number of swarm participants, starting relative coordinates, and statuses (imager and/or strike), and drives the establishment of critical swarm and target jobs. The server uses complex algorithms, such as the research road graph, which is based on the rotor-router model, and the complete information exchange graph, which uses the gossip/broadcast model, to carry out these tasks. These algorithms complement each other in the server environment, allowing the UAV swarm to plan and coordinate tasks more effectively. In addition, the platform provides the smooth transfer of the created goal tasks to each member of the swarms’ memory, benefiting decision-making skills and swarm performance as an entire. The QT-based user interface also provides the capability to modify the specific member task type of UAVs. This implies that UAVs can be used for specific area consideration, information gathering, transport, or, in the case of a military use, as an enemy position attacker. A server environment and QT-based interface can be used as a data sharing transponder to accomplish all of these functional changes. The task categories within the mentioned swarm shift when alterations occur in the initial transfer graph of UAVs, necessitating the inclusion of new locations for completing the modified tasks. Additionally, ensuring a unified primary administrative domain and coordinated data transmission among users across multiple platforms are crucial aspects of the suggested method. This domain ensures secure connections and offers additional security features.

The problem of passive localization using an unmanned aerial vehicle (UAV) swarm is studied. For multi-UAV localization systems with limited communication and observation range, the challenge is how to obtain accurate target state consistency estimates through local UAV communication. In this paper, an ant colony pheromone mechanism-based passive localization method using a UAV swarm is proposed. Different from traditional distributed fusion localization algorithms, the proposed method makes use of local interactions among individuals to process the observation data with UAVs, which greatly reduces the cost of the system. First, the UAVs that have detected the radiation source target estimate the rough target position based on the pseudo-linear estimation (PLE). Then, the ant colony pheromone mechanism is introduced to further improve localization accuracy. The ant colony pheromone mechanism consists of two stages: pheromone injection and pheromone transmission. In the pheromone injection mechanism, each UAV uses the maximum likelihood (ML) algorithm with the current observed target bearing information to correct the initial target position estimate. Then, the UAV swarm weights and fuses the target position information between individuals based on the pheromone transmission mechanism. Numerical results demonstrate that the accuracy of the proposed method is better than that of traditional localization algorithms and close to the Cramer–Rao lower bound (CRLB) for small measurement noise.

The resilience of unmanned aerial vehicle (UAV) swarm systems describes the capability to adjust its behavior to maintain its essential functions when malfunction or attack occurs. Recent studies have modeled the UAV swarm as a network to analyze its resilience, where each UAV is represented as a node, and each link denotes information exchange between UAVs. However, the current networking method is vulnerable to malicious attacks and does not consider the individual load of one UAV is limited. In this paper, we proposed a resilient UAV swarm networking model by incorporating the effect of the limitation of the bandwidth. Under continuous malicious attacks, this proposed model could enhance the resilience of the swarm systems by rewiring the nodes. Moreover, a case study is conducted in which a UAV swarm is to perform a search and rescue task. Experimental results show that the proposed model is more resilient to malicious attacks than some existing models, e.g., the BA model, ER model, and the balanced individual communication load.

Unmanned aerial vehicle swarm distributed cooperation method based on situation awareness consensus and its information processing mechanism

This paper develops the model of the joint phantom track deception and the joint techniques of time-difference of arrival (TDOA) and frequency-difference of arrival (FDOA) localization to deceive air defense radar networks under the condition that an unmanned aerial vehicle (UAV) swarm has no prior knowledge of the radars’ precise locations, and related performance experiment and analysis are presented to demonstrate the effectiveness of the proposed method and to clarify the influence factors of phantom track deception. The main contributions of this paper are as follows. Firstly, the model of phantom track deception against a radar network by UAV swarm without prior knowledge of the radars’ positions are established. Secondly, TDOA/FDOA are adapted to locate networked enemy radars using UAV swarm, where the Fisher information matrix (FIM) is derived to evaluate the estimation accuracy. Thirdly, the uncertainty analysis consisting of radar location error and UAV position error is deduced. With these efforts, the integrated capability of sensing and jamming is realized. Moreover, the same source testing using space resolution cell (SRC) from the perspective of a radar network is executed to provide guidance for phantom track design. Finally, performance experiment and analysis are given to verify the theoretical analysis with simulation results.

In recent years, the concept of UAV (Unmanned Aerial Vehicle) swarm has been proposed and developed, which effectively solves the shortcomings of relatively small payload and relatively weak information perception processing capability in a single UAV operation. The research and application of UAV swarm technology has become an important direction of the development of UAV technology. UAV swarm not only can effectively improve the load capacity and information processing capacity through close cooperation between single UAV, but also has a high "self-healing" ability and strong robustness. However, due to the limited computing resources and complex external environment of UAVs, UAV swarm are vulnerable to forgery attacks, man-in-the-middle attacks and reply attacks. It is necessary to establish an authentication scheme to ensure the legitimate and reliable identity of UAVs for data exchange and sharing. However, traditional authentication schemes based on username, password or dynamic key have lower security levels, while RSA authentication requires a long session key and cannot meet the lightweight requirements in an UAV swarm. Based on the analysis and summary of existing authentication technologies and the special requirements of UAV swarm for cyber security, this paper puts forward an authentication scheme based on Zero Trust, achieves rapid authentication of UAV swarm in data exchange and sharing, and strengthens its ability to respond to cyber attacks.

Mission reliability evaluation of high‐performance unmanned aerial vehicle (UAV) swarms faces significant challenges because of their tightly coupled mission profiles, dynamic reconfiguration mechanisms, and heterogeneous node interactions. Traditional reliability models fail to capture these complexities, particularly the impact of mission reconstruction on the UAV swarm adaptability. To address this gap, we propose a novel mission reliability evaluation framework that integrates an effective mission loop analysis and dynamic network reconfiguration. First, we analyzed the system composition and interaction of high‐performance UAV swarms and designed a mission reliability evaluation model based on effective mission loops. Second, we developed a dynamic reconstruction algorithm to autonomously reconfigure the swarm topology in response to node/edge failures, ensuring sustained operational capability. Third, by integrating the reliability evaluation model with the reconstruction algorithm, we built a high‐performance UAV swarm mission reliability evaluation method that explicitly incorporates mission reconstruction. A case study of a 10‐UAV swarm strike mission validated the proposed framework. The results demonstrate its effectiveness in mission reliability evaluation of high‐performance UAV swarms, with potential extensions to heterogeneous aviation systems.

Unmanned aerial vehicle (UAV) swarm networks have various applications. They support surveillance and communication. However, these ‎networks face many challenges. High mobility causes frequent link breakages. Limited transmission range increases interference. Routing ‎becomes complex due to these factors. This paper proposes a new approach. It jointly optimizes trajectory control, frequency allocation and ‎routing. The proposed algorithm is called JTFR. It improves link utility in UAV swarm networks. Link utility depends on signal strength, ‎queue delay and residual energy. The proposed JTFR uses multi-agent reinforcement learning. The learning model considers both local and ‎neighbour information. This improves the stability of the network. A long short-term memory (LSTM) network is used. This helps in pre-‎dicting the best trajectory. A multi-head attention mechanism is also used. Each UAV adjusts its policy based on its neighbours. UAV ‎swarm networks are dynamic. The movement of UAVs changes frequently. This affects network connectivity. The routing decisions must ‎adapt to these changes. The proposed approach provides adaptive routing. JTFR is overall useful to improve UAV swarm communication. ‎It handles trajectory control, frequency allocation and routing in balancing. The method is dynamic in the sense that it adapts to dynamic ‎network conditions. The UAVs work well in complex environments‎.

Task allocation for Unmanned Aerial Vehicle (UAV) swarms is a complex challenge. Effective distribution of tasks among multiple UAVs necessitates the optimization of conflicting objectives, such as task balance, flight distance, cost, and target benefits. This multi-objective optimization problem is inherently intricate due to strong interdependencies, nonlinear relationships, and dynamic environments. Traditional methods, including heuristic algorithms, supervised machine learning, and basic reinforcement learning mechanisms, often suffer from low efficiency and strict requirements for data sources or supervision mechanisms. They also struggle to adapt to problems with large discrete action spaces. To address this challenge, this paper proposes a novel approach that integrates swarm intelligence simulation with reinforcement learning. The core of this approach is a reinforcement learning algorithm based on the Deep Deterministic Policy Gradient (DDPG) framework. A key innovation is a customized DDPG variant specifically designed to handle the large, discrete action spaces inherent in UAV task allocation. Additionally, a behavior tree model is employed to simulate UAV swarm interactions, providing a realistic and verifiable environment for evaluating task allocation strategies. This simulation generates a rich dataset for training and testing reinforcement learning algorithms. The combined approach efficiently addresses the multi-objective optimization problem, enabling optimal task distribution for numerous UAVs. Compared to existing heuristic methods, the proposed approach demonstrates superior computational efficiency and solution quality, while maintaining reasonable convergence stability and environmental adaptability. Simulation-based validation ensures the approach’s robustness in dynamic environments, a critical factor for practical applications. Overall, the integration of simulation and reinforcement learning provides a powerful framework for tackling the complex task allocation problem in UAV swarms.

This paper is concerned with a new data-driven zero-sum neuro-optimal control problem for continuous-time unknown nonlinear systems with disturbance. According to the input-output data of the nonlinear system, an effective recurrent neural network is introduced to reconstruct the dynamics of the nonlinear system. Considering the system disturbance as a control input, a two-player zero-sum optimal control problem is established. Adaptive dynamic programming (ADP) is developed to obtain the optimal control under the worst case of the disturbance. Three single-layer neural networks, including one critic and two action networks, are employed to approximate the performance index function, the optimal control law, and the disturbance, respectively, for facilitating the implementation of the ADP method. Convergence properties of the ADP method are developed to show that the system state will converge to a finite neighborhood of the equilibrium. The weight matrices of the critic and the two action networks are also convergent to finite neighborhoods of their optimal ones. Finally, the simulation results will show the effectiveness of the developed data-driven ADP methods.

Adaptive decision-making with deep Q-network for heterogeneous unmanned aerial vehicle swarms in dynamic environments

This paper investigates the issue of spectrum collaboration under dynamic changes in the number of unmanned aerial vehicle (UAV) swarms under jamming conditions. We focus on the scenario where multiple UAV swarms conduct reconnaissance tasks, the head UAV (H-UAV) of each swarm relays the detected information gathered by the swarm back to the ground control station (GCS). To achieve spectrum collaboration among H-UAVs, first, we model the problem as an optimization that maximizes the sum of the achievable rates of H-UAVs to GCS. Then, we treat H-UAVs as agents and adopt the classic multi-agent proximal policy optimization (MAPPO) for centralized training of the agents. By setting the number of agents to a specific value, we make it applicable for distributed spectrum collaboration scenarios with dynamic changes in the number of UAV swarms. Finally, we measure the effectiveness of the algorithm by the spectrum conflict occurrences. Simulation results show that the proposed algorithm can significantly reduce the number of spectrum conflicts among H-UAVs, achieving effective spectrum collaboration under dynamic changes in the number of UAV swarms in jamming conditions.

Introduction: Unmanned aerial vehicle (UAV) swarms are increasingly employed in civilian and military applications. Efficient path planning is critical for navigating dynamic environments, yet challenges such as collision avoidance, energy optimization, and multiagent coordination hinder deployment. Methods: A systematic literature review (SLR) was conducted to classify UAV swarm pathplanning algorithms into four categories: sampling-based, graph-based, biologically inspired, and mathematical model-based approaches. The review adhered to established SLR guidelines, employing systematic database searches and predefined inclusion and exclusion criteria. Each algorithm was evaluated in terms of computational efficiency, adaptability, and application suitability. Results: Biologically inspired algorithms demonstrated high adaptability and flexibility, whereas graph-based approaches were effective in structured environments. Sampling-based methods proved scalable in high-dimensional spaces, and mathematical models emphasized optimization. Persistent challenges include real-time adaptability, energy efficiency, and operations in interactive environments. Discussion: The comparative analysis highlights the trade-offs inherent in each approach and underscores the unresolved challenges that limit practical implementation. Addressing these gaps is essential to advance the reliability and applicability of UAV swarms in real-world scenarios. Conclusion: This review systematically evaluates UAV swarm path-planning algorithms, identifying their respective advantages, limitations, and application contexts. The findings provide a reference framework to inform future research directions aimed at enhancing UAV swarm autonomy and operational efficiency.