Underwater exploration activities have grown significantly due to the proliferation of underwater Internet of Things (UIoT). However, to transmit sensor data from UIoT to remote onshore data processing center requires huge cost of deploying and maintaining communication infrastructures. In this paper, we propose an Unmanned Aerial Vehicles (UAVs)-assisted underwater data acquisition scheme by placing multiple sink nodes on the water surface to serve as intermediate relays between underwater sensors (IoT nodes) and UAVs. In our scheme, the sensor data is first transmitted via an acoustic-signal link to a buoyant sink node, which then forwards the data to a UAV via an electromagnetic link. In particular, we adopt two sink-node-deployment methods, i.e., grid placement and random placement of sink nodes. Since the path connectivity from an underwater sensor node to the UAV is crucial to guarantee reliable data acquisition tasks, we establish a theoretical framework to analyze the path connectivity via the intermediate sink node for both grid and random sink-node-deployment methods. Extensive simulation results validate the accuracy of the proposed analytical model. Moreover, our results also reveal the relationship between the path connectivity and other factors, such as sink node placements, antenna beamwidth of UAVs and wind speed. We also further extend our UAV-assisted data acquisition to other scenarios with consideration of trajectories of UAVs, movements of sink nodes, interference of both underwater acoustic and terrestrial radio links and integration with edge computing.

In the 6G era, the proliferation of data and data-intensive applications poses unprecedented challenges on the current communication and computing networks. The collaboration among cloud computing, edge computing, and networking is imperative to process such massive data, eventually realizing ubiquitous computing and intelligence. In this article, we propose a wireless computing power network (WCPN) by orchestrating the computing and networking resources of heterogeneous nodes toward specific computing tasks. To enable intelligent service in WCPN, we design a task and resource-aware federated learning model, coined FedTAR, which minimizes the sum energy consumption of all computing nodes by the joint optimization of the computing strategies of individual computing nodes and their collaborative learning strategy. Based on the solution of the optimization problem, the neural network depth of computing nodes and the collaboration frequency among nodes are adjustable according to specific computing task requirements and resource constraints. To further adapt to heterogeneous computing nodes, we then propose an energy-efficient asynchronous aggregation algorithm for FedTAR, which accelerates the convergence speed of federated learning in WCPN. Numerical results show that the proposed scheme outperforms the existing studies in terms of learning accuracy, convergence rate, and energy saving.

By integrating the merits of aerial, terrestrial, and satellite communications, the space-air-ground integrated network (SAGIN) is an emerging solution that can provide massive access, seamless coverage, and reliable transmissions for global-range applications. In SAGINs, the uplink connectivity from ground users (GUs) to the satellite is essential because it ensures global-range data collections and interactions, thereby paving the technical foundation for practical implementations of SAGINs. In this article, we aim to establish an accurate analytical model for the uplink connectivity of SAGINs in consideration of the global distributions of both GUs and aerial vehicles (AVs). Particularly, we investigate the uplink path connectivity of SAGINs, which refers to the probability of establishing the end-to-end path from GUs to the satellite with or without AV relays. However, such an investigation on SAGINs is challenging because all GUs and AVs are approximately distributed on a spherical surface (instead of the horizontal surface), resulting in the complexity of network modeling. To address this challenge, this paper presents a new analytical approach based on spherical stochastic geometry. Based on this approach, we derive the analytical expression of the path connectivity in SAGINs. Extensive simulations confirm the accuracy of the analytical model.

The spotlight on autonomous aerial vehicles (AAVs) is to enhance wireless communications while ignoring the potential risk of AAVs acting as adversaries. Due to their mobility and flexibility, AAV eavesdroppers pose an immeasurable threat to legitimate wireless transmissions. However, the existing fixed jamming scheme without cooperation cannot counter the flexible and dynamic AAV eavesdropping. In this article, a cooperative intelligent jamming scheme is proposed, authorizing ground jammers (GJs) to interfere with AAV eavesdroppers, generating specific jamming shields between AAV eavesdroppers and legitimate users. Toward this end, we formulate a secrecy capacity maximization problem and model the problem as a decentralized partially observable Markov decision process (Dec-POMDP). To address the challenge of the huge state space and action space with network dynamics, we leverage a deep reinforcement learning (DRL) algorithm with a dueling network and double-Q learning (i.e., dueling double deep Q-network) to train policy networks. Then, we propose a multi-agent mixing network framework (QMIX)-based collaborative jamming algorithm to enable GJs to independently make decisions without sharing local information. Additionally, we perform extensive simulations to validate the superiority of our proposed scheme and present useful insights into practical implementation by elucidating the relationship between the deployment settings of GJs and the instantaneous secrecy capacity.

With the help of artificial intelligence (AI) model, aerial computing can help analyze and predict the network dynamics and support intelligent decision-making to improve the performance of 6G space-air-ground integrated networks. Federated learning has been proposed to tackle the challenges of limited energy and data shortage for the application of AI models in aerial computing networks. A critical problem of FL for aerial computing is the lack of incentives due to privacy concerns. On the one hand, the information needed to measure users' learning quality may be eavesdropped. On the other hand, users' real costs for determining payments may also undertake inference attacks. In this paper, we design a privacy-preserving and learning quality-aware incentive mechanism for federated learning in aerial computing networks. We propose differential privacy based scheme to protect the privacy of the real cost. In addition, utilize Combinatorial Multi-Armed Bandit (CMAB) algorithm to evaluate the user learning quality without any participant information. Simulation results demonstrate that our scheme can significantly motivate high-quality participants with guaranteed privacy preservation and achieve effective federated learning under the constraint of the limited budget.

Wireless Computing Power Networks (WCPN), guided by green principles, aim to provide efficient, flexible, and environmentally friendly computing services for Internet of Things (IoT) applications by seamlessly coordinating computational and networking resources across diverse nodes. The integration of Digital Twin (DT) technology is crucial for achieving these objectives. However, different DT association strategies play a crucial role in enhancing the capabilities of WCPN. In this paper, recognizing the long-term and dynamic nature of DT deployment in real-world scenarios, we utilize evolutionary game theory to model the association and transfer of DTs, aiming for continuous adaptive adjustments and optimizations in their deployment. Specifically, we propose an evolutionary game-based algorithm for DT association as a complement to the independent decision-making process in DT deployment. Moreover, in light of the inherent limitations of the evolutionary game selection mechanism and the lack of self-learning ability, we introduce a deep Q-network (DQN) based evolutionary game approach that ensures adaptive DT association and transfer by considering factors such as DT synchronization delay, model consistency, and migration costs. Numerical results demonstrate that our proposed algorithms outperform the benchmarks in terms of average user utility and convergence speed.