Data collection and distributed task execution in Internet of Things (IoT) networks require efficient coordination among autonomous agents to handle the growing volume of sensing and computational demands. Unmanned aerial vehicles (UAVs) and unmanned ground vehicles (UGVs) present promising candidates for these operations due to their complementary capabilities and mobility advantages. However, effective cooperation between these heterogeneous agents faces significant challenges including communication limitations, energy constraints, and suboptimal task allocation efficiency. In this paper, we aim to maximize data collection capacity, task completion rates, while minimizing energy consumption across all UAVs. We propose U2GNet, a novel UGV-assisted framework for UAV that enables efficient task offloading and resource allocation in dynamic environments by leveraging Deep Reinforcement Learning (DRL) enhanced with Heterogeneous Graph Attention Networks (HGAT). The framework employs HGAT to process local observations and information shared by neighboring agents, while Gated Recurrent Units (GRU) address partial observability by integrating historical information, and DRL optimizes the decision-making process. Simulation results demonstrate that U2GNet improves the average data collection rate and task completion rate by 16.90% and 10.81% respectively compared to the baseline HGN approach.

Forward-looking sonar (FLS) image segmentation can help reduce the amount of raw data that needs to be transmitted in underwater communication systems, making it a crucial technique for next-generation communication systems and the Internet of Things (IoT). However, its effectiveness is often hindered by weak semantic information, blurry edges and low resolution, which pose challenges for current segmentation algorithms. In this study, we propose a multimodal feature-enhanced Unet for FLS image segmentation (MFEUnet), built upon the Unet framework. The multimodal features considered primarily include spatial and frequency features. For spatial features, recognizing Unet’s strength in local feature extraction, we integrate a transformer to enhance its ability to capture global features. Frequency features are utilized to capture different details of FLS images, with a dual-branch wavelet transformation employed to decompose images into low-frequency and high-frequency components, facilitating the enhancement of these features. And a preprocessing reconstruction module is integrated to reduce the noise of FLS images. Furthermore, to address class imbalance in FLS datasets, we design a specialized segmentation loss function. Experimental results show that MFEUnet significantly outperforms state-of-the-art segmentation methods, demonstrating its effectiveness in overcoming the unique challenges of underwater sonar imaging.

Mobile crowd sensing (MCS) is an emerging paradigm that enables participants to collaborate on various sensing tasks. UAVs are increasingly integrated into MCS systems to provide more reliable, accurate and cost-effective sensing services. However, optimizing UAV trajectories and communication efficiency, especially under non-line-of-sight (NLoS) channel conditions, remains a significant challenge. This paper proposes TRAIL, a Transformer-enhanced deep reinforcement Learning (DRL) algorithm. TRAIL aims to jointly optimize UAV trajectories and Reconfigurable Intelligent Surface (RIS) phase shifts to maximize data throughput while minimizing UAV energy consumption. The optimization problem is modeled as a Markov Decision Process (MDP), where the Transformer architecture captures long-term dependencies in UAV trajectories, and these features are input into a Double Deep Q-Network with Prioritized Experience Replay (PER-DDQN) to guide the agent in learning the optimal strategy. Simulation results demonstrate that TRAIL significantly outperforms state-of-the-art methods in both data throughput and energy efficiency.

Aerial computing paradigms, particularly those involving UAVs as access points and radio towers, show significant promise for local data analysis and real-time service provision in aerial access networks. However, the limited battery life of UAVs, compounded by the high-energy demands of communication tasks and prolonged computation delays, presents a significant challenge. Deep reinforcement learning (DRL) enables UAVs to autonomously optimize their operations, reducing both energy consumption and latency. Nevertheless, the prolonged learning process of DRL can lead to inefficiencies, especially in dynamic environments where the equitable participation of UAVs is crucial. To address these issues, we introduce a fairness-oriented federated learning (FL) scheme that employs importance sampling to select UAVs for training, ensuring equitable utilization of each UAV's data. Furthermore, we integrate this FL fairness scheme into the design of DRL algorithms, termed FedDRL. This algorithm jointly optimizes the computation capabilities and bandwidth allocation of UAVs to minimize system costs. Numerical results demonstrate the fairness of FedDRL in fFL networks. Specifically, compared to other state-of-the-art DRL algorithms (e.g., TD3 and DDPG), FedDRL reduces system costs by 43.82% and 49.45%, respectively.

With the advancements in technologies such as 5G, Unmanned Aerial Vehicles (UAVs) have exhibited their potential in various application scenarios, including wireless coverage, search operations, and disaster response. In this paper, we consider the utilization of a group of UAVs as aerial base stations (BS) to collect data from IoT sensor devices. The objective is to maximize the volume of collected data while simultaneously enhancing the geographical fairness among these points of interest, all within the constraints of limited energy resources. Therefore, we propose a deep reinforcement learning (DRL) method based on Graph Attention Networks (GAT), referred to as “GADRL”. GADRL utilizes graph convolutional neural networks to extract spatial correlations among multiple UAVs and makes decisions in a distributed manner under the guidance of DRL. Furthermore, we employ Long Short-Term Memory to establish memory units for storing and utilizing historical information. Numerical results demonstrate that GADRL consistently outperforms four baseline methods, validating its computational efficiency.