Search

Unmanned aerial vehicles (UAVs) deployed as a swarm can offer a flexible and cost-efficient solution for surveillance missions in large-scale adversarial environments. UAV swarms possess superior real-time surveillance capabilities in detecting time-sensitive targets with speed and security, surpassing alternative techniques such as satellites and ground-based sensors. However, the area surveillance problem in adversarial scenarios using UAV swarms necessitates careful consideration of factors such as swarm positioning, flight attitudes, antenna directionality, the threat posed by adversary detection systems (ADSs), and swarm network topology. To address this challenge, we propose the Integrated Low Detection Probability (ILDP) deployment method for UAV swarms, which incorporates the Danger Avoidance Distributed Motion control algorithm (DADM) for swarm collaboration and the Low Detection Probability Topology Control algorithm (LDPTC) for swarm networking. The DADM algorithm facilitates swarm cooperation in achieving area coverage and evading ADSs by leveraging neighboring, environmental, and threat information in adversarial scenarios. Furthermore, the LDPTC algorithm establishes a topology optimization model that comprehensively considers swarm distribution and the impact of directional antenna sidelobes to reduce transmitting energy leakage on the ground. Our strategy significantly decreases the detection probability of UAV swarms by ADSs, ensuring the operational effectiveness of UAV swarms in dynamic adversarial scenarios. Extensive simulations validate the superiority of our proposed ILDP method, demonstrating considerably lower detection probabilities compared to other approaches in static and dynamic adversary environments across various swarm scales. Moreover, our method excels in real-time surveillance capabilities with lower computational complexity, freeing up computing resources for UAVs to fulfill additional tasks.

This paper considers the cooperative search for stationary targets by multiple unmanned aerial vehicles (UAVs) with limited sensing range and communication ability in a dynamic threatening environment. The main purpose is to use multiple UAVs to find more unknown targets as soon as possible, increase the coverage rate of the mission area, and more importantly, guide UAVs away from threats. However, traditional search methods are mostly unscalable and perform poorly in dynamic environments. A new multi-agent deep reinforcement learning (MADRL) method, DNQMIX, is proposed in this study to solve the multi-UAV cooperative target search (MCTS) problem. The reward function is also newly designed for the MCTS problem to guide UAVs to explore and exploit the environment information more efficiently. Moreover, this paper proposes a digital twin (DT) driven training framework “centralized training, decentralized execution, and continuous evolution” (CTDECE). It can facilitate the continuous evolution of MADRL models and solve the tradeoff between training speed and environment fidelity when MADRL is applied to real-world multi-UAV systems. Simulation results show that DNQMIX outperforms state-of-art methods in terms of search rate and coverage rate.