Due to their high flexibility and strong maneuverability, unmanned aerial vehicles (UAVs) have attracted lots of attention and are widely employed in many fields. Especially in target encirclement applications, UAVs have shown great advantages in adaptability and reliability, and can efficiently fly to and evenly surround the targets in complex and dynamic environments. In this paper, we concentrate on the cooperative target encirclement problem of heterogeneous UAVs and try to propose a multi-agent reinforcement learning approach to solve the problem. First, with the models of heterogeneous UAVs and obstacles, we analyze the collision avoidance, motion continuity, and energy consumption constraints of UAVs, and formulate the cooperative target encirclement problem as a multi-constraint combinatorial optimization one. Then, inspired by the humans’ learning experience that curiosity provides a powerful motivator for humans to explore, discover, and acquire new knowledge, we propose an extrinsic-and-intrinsic reward-based multi-agent reinforcement learning approach to cooperatively control the behaviors of UAVs and achieve the target encirclement missions. Simulation experiments with randomly generated environments are conducted to evaluate the performance of our approach, and the results show that our approach has a significant advantage in terms of average reward, encirclement success rate, encirclement time, and encirclement energy consumption.

Collaborative perception effectively mitigates issues such as limited field of view and occlusion by enabling multiple agents to share perceptual information. Despite its advantages, challenges persist in complex environments due to factors such as limited communication bandwidth and noisy poses, which may potentially degrade system performance. Meanwhile, a substantial amount of simulation data is widely adopted in collaborative perception to achieve high precision and real-time detection. However, the domain gap between simulated and real-world environments may result in weakened collaborative performance and hindered generalization ability. In this work, we focus on the multi-agent collaborative perception problem and propose a quantized communication and transferable fusion framework, named QCTF, to efficiently minimize the bandwidth overhead and enhance real-world perception by leveraging unlabeled data for improved adaptability. First, we present a quantized communication method that employs multi-scale residual indices and an optimized codebook to extract robust representations while minimizing bandwidth usage. Then, we design a channel-aware selection strategy that adjusts the bandwidth volume and compensates for the quantized representation by combining the prioritized critical features with the channel dimension. Finally, we adopt a transferable fusion module to effectively bridge the simulation-to-reality domain gaps and improve perceptual capability through multi-scale adaptation discriminators. Experiments on both simulated and real-world datasets are conducted to evaluate the effectiveness of the proposed framework, and the results demonstrate that our approach consistently outperforms the existing methods in limited communication bandwidth and domain adaptation scenarios.

Unmanned aerial vehicles (UAVs) have been extensively adopted in mobile edge computing (MEC) due to their high autonomy and powerful maneuverability. Although UAV-assisted MEC can achieve strong efficiency and flexibility enhancements for mobile users, it results in serious resource allocation and task scheduling problems. The computing, storage, and communication resources of UAVs should be efficiently allocated and scheduled so that the service quality at different positions is ensured even in a highly dynamic environment. In this work, we focus on the UAV-assisted MEC problem and design a pointer network-based reinforcement learning framework that effectively addresses the pain point of poor reusability of traditional reinforcement learning methods and optimizes the decision-making behavior of UAVs to improve the system performance. First, with the models of UAVs and mobile users, we analyze the collision avoidance, motion continuity, coverage, and energy consumption constraints of UAVs, and formulate the UAV-assisted MEC problem as a multi-constraint combinatorial optimization one. Then, inspired by the suitability of pointer networks in effectively handling variable-length sequences, we propose a pointer network-based reinforcement learning framework to seek an optimal service strategy for every UAV while minimizing the completion time of MEC tasks. Simulation experiments are conducted on randomly generated environments to quantitatively analyze the performance of our approach, and the experimental results indicate that compared to traditional methods, our approach can reduce the task completion time, flight path length, and energy consumption by more than 1.25%, 2.78%, and 1.12%, respectively.

Collaborative perception facilitates information exchange and communication among neighbouring agents and is a competitive solution to address the limited field of view of a single vehicle in autonomous driving. Although collaborative perception has brightened application prospects and can greatly enhance vehicle safety in transportation systems, it often requires substantial communication and computation overhead to achieve efficient collaboration. Even worse, the sparse point cloud data caused by distant objects impairs the perceptual abilities of individual agents, and unavoidable errors in agents' poses result in misaligned global observations and low perception accuracy. In this work, we focus on the multi-agent collaborative perception problem and present a Collaborative Range-aware adaptive fusion framework named CoRange to achieve the communication-efficient and fusion-effective perception of the rapidly changing environment. First, we present a range-aware communication mechanism to reduce the misjudgment and minimize the bandwidth consumption, by fully considering the characters of range information and selectively transmitting the features of critical regions. Then, we design a local and agent-wise attention module to handle point clouds at different distances and capture relationships between heterogeneous agents, with an objective of enhancing the detection accuracy and robustness against pose errors. Finally, we adopt a hierarchical adaptive fusion method to effectively integrate the features representing diverse semantic information and provide multi-source representations for ego agents. Experiments on both simulated and real-world datasets are conducted to validate the efficiency and effectiveness of the proposed approach, and the experimental results demonstrate that our approach exhibits superior performance in limited communication bandwidths and noisy environments when LiDAR-based object detection tasks are carried out.

Collaborative perception enables multiple agents to exchange and integrate information, achieving a more comprehensive understanding of the surrounding environment. While it holds great potential to mitigate object occlusion and compensate for individual agents’ limited field of view, several challenges remain. Among them are localization noise and transmission delays in complex environments, which may cause spatial and temporal misalignment of shared features. Additionally, limited communication bandwidth hinders the effective utilization of collaborative information, resulting in suboptimal feature aggregation. To tackle these challenges, we present CEST, a Communication-Efficient Spatial-Temporal fusion framework that jointly considers historical context and spatial misalignment while reducing communication overhead. First, we present a temporal context aggregation module that encodes temporal information and captures collaborators’ temporal context to enhance current feature estimation. Then, we adopt a query-aware communication mechanism that leverages both channel and spatial queries to identify visually essential and complementary information, while incorporating a mutual information loss to preserve discriminative features. Finally, we design a spatial-temporal aggregation module that enables efficient fusion by integrating multi-scale spatial representations and extracting contextual cues across agents. To thoroughly assess our framework, we conduct experiments on both simulated and real-world collaborative 3D object detection datasets. The results confirm that our approach maintains robust performance across varying conditions, including pose errors, transmission delays, and bandwidth limitations.