With the wide application of unmanned aerial vehicles (UAVs), performing search and rescue missions autonomously in unknown environment has become an increasingly concerning issue. In this paper, we propose an adaptive conversion speed Q-Learning algorithm (ACSQL). Performing UAV missions autonomously is divided into two stages: rescue mission search stage and optimal path search stage. In the first stage, a UAV can find task points as soon as possible, and the efficiency of exploration is increased by adaptively adjusting the speed of the UAV. In the second stage, to get a secure and short path, we propose a subdomain search algorithm. Based on the above two stages, we improve state space and action space in reinforcement learning, and design a composite reward function, finally obtain the path of UAV to perform multiple search and rescue missions through this algorithm. In order to solve the problems of slow training convergence and high uncertainty, we initialize the Q-table by combining detection information of UAV sensors in first stage. Simulation results show that ACSQL algorithm can realize autonomous navigation and path planning of UAV in an unknown environment. Compared with traditional action space, the learning process of UAV converges faster and more stable, and it can converge in about 30 episodes. Compared with DDPG algorithm and IDWA algorithm in different scenarios, ACSQL algorithm has the shortest path length. Finally, ACSQL algorithm is verified by UAV simulator Airsim.

The problem of task allocation of UAV systems is a hot issue in scientific research. Aiming at overcome the shortcomings of existing algorithms in terms of load balancing and execution efficiency, an algorithm Improved Self-Organizing Mapping (ISOM) is proposed in this paper. Firstly, considering the two factors of the UAV flight distance and the mission execution time, the load balance of the UAV is designed. The sum of the track lengths of multiple UAVs and the total time required by multiple UAVs to complete all the tasks in the task area are used to evaluate the quality of cooperative task assignment of multiple UAVs. Secondly, the learning rate and neighborhood function of nonlinear change are designed to ensure the stability and accelerate its convergence. With the increase of the number of iterations, the radius of the neighborhood gradually decreases, and the renewal range of the superior neighborhood is determined by the nonlinear function. Finally, in order to solve the problem of new tasks of UAV and UAV failure, the attention mechanism based on the ISOM algorithm is introduced. Different attention is allocated by the distance between the UAV and the mission points to ensure the missions can be fully executed. Compared with algorithms PSO-GA, Gurobi and ORTools, the time to complete the task is effectively reduced, respectively. At the same time, the algorithm is verified in a large task environment. When the number of tasks is 200 and the number of UAVs is 8, in the TSPLIB task set, the simulation results demonstrate the high efficiency and flexibility of the proposed algorithm.

With the further development of Unmanned Aerial Vehicle (UAV) technologies, research on multi-UAV formations have also received more attention. Unmanned Aerial Vehicles (UAVs) cooperate with each other to form a formation group, which can give full play to the advantages that a single UAV does not have, and more capable of working in multi-task scenarios. Based on the Sierpinski fractal structure, this paper proposes a hybrid formation control architecture. To address the aggregation problem of UAV swarms, a multi-agent deep reinforcement learning (MADRL) method is used for aggregation control. To improve the efficiency, accuracy, effectiveness and scalability of MADRL in environments with a large number of UAVs, a multi-intelligent deep deterministic policy gradient method based on attention mechanism and adaptive accuracy (3A-MADDPG) is proposed. The method enables each agent to selectively focus on the information of other agents, with adaptive learning rate to dynamically learn its own critic network. The algorithm has a large learning rate in the early stage and converges quickly. The later stage of the algorithm has small learning rate and high accuracy, and combined with the fixed variable hybrid reward function designed in the paper, so the UAV can fly to the center of the expected region more accurately, making the whole cluster more accurate. Experiments show that the 3A algorithm proposed in the paper converges faster and achieves higher accuracy than the benchmark algorithm, whether it is the cluster aggregation formation or the separation and aggregation of sub-formations.

With the proliferation of machine learning models in diverse applications, the issue of model security has increasingly become a focal point. Model steal attacks can cause significant financial losses to model owners and potentially threaten the security of their application scenarios. Traditional model steal attacks are primarily directed at soft-label black boxes, but their effectiveness significantly diminishes or even fails in hard-label scenarios. To address this, for hard-label black boxes, this study proposes an active learning-based Fusion Augmentation Model Stealing Framework (AugSteal). This framework initially utilizes large-scale irrelevant public datasets for deep filtering and feature extraction to generate reliable, diverse, and representative high-quality data subsets as the stealing dataset. Subsequently, we developed an adaptive active learning selection strategy that selects data samples with significant information gain for different black-box models, enhancing the attack’s specificity and effectiveness. Finally, to further address the trade-off between query budget and steal precision, this paper designed a Fusion Augmentation training method constituted of two different loss functions, enabling the substitute model to closely approximate the decision distribution of the target black box.The comprehensive experimental results indicate that, compared to the current state-of-the-art attack methods, our approach achieved a maximum performance gain of 8.21% in functional similarity for the substitute models in simulated black-box scenarios CIFAR10, SVHN, CALTECH256, and the real-world application Tencent Cloud API.

FANET (Fly-Adhoc-Network) does not rely on prebuilt infrastructure, and can form a temporary network through wireless links anytime and anywhere, which has been widely used in emergency communication and disaster relief. In order to solve the problem of signal transmission fading caused by heavy fog, heavy smoke and other harsh environments, and data packet transmission failure caused by high speed movement of unmanned aerial vehicle (UAV), a cooperative transmission mode is adopted to select a relay at the data link layer for packet forwarding, which effectively improves the network communication performance. Aiming at the problem of limited communication of UAV swarm in harsh environment, a cooperative medium access control protocol named energy consumption and channel gain cooperative medium access control protocol (EC-CMAC) for FANET is proposed. The protocol estimate the transmit power according to the channel propagation model, and further propose a relay selection strategy according to the estimate transmit power, the residual energy of the node, the direction and position of the node, and adaptively selecting the transmission mode. During cooperative transmission, it selected one-hop neighbors to forward the message, and extended the network lifetime by optimizing the energy consumption of the UAVs. The protocol is verified both statically and dynamically in MATLAB environment. The evaluation results of end-toend delay, network throughput, network lifetime and packet transmission ratio testify that EC-CMAC can achieve longer network lifetime and higher packet delivery rate with little delay and throughput loss in harsh transmission environment and high speed movement of UAVs.