Multi-beam satellites (MBS) exploit multiple spot beams and frequency reuse to achieve high spectral efficiency and flexible coverage, which are key characteristics of modern satellite communication systems. Among them, beam hopping (BH) leverages phased-array antennas to steer onboard beams and employs time-division multiplexing (TDM) to dynamically schedule illumination patterns in the time domain, with electronic beam position switching enabling rapid reconfiguration. However, existing works exhibit two critical limitations: (1) they rely on global decision strategies that incur high inference complexity, which further increases with the number of beams and ground cells, making it difficult to balance performance and computational complexity; (2) in pursuing high spectral reuse, they overlook beam overlap and co-frequency interference (CFI), and thus cannot effectively mitigate interference without sacrificing spectral efficiency. To address these challenges, this paper proposes a cell clustering-based BH (CCBH) algorithm with low complexity and interference avoidance. Specifically, we propose a region-growing cell-clustering method based on user demand load balancing in which each beam independently serves a cell cluster, and full-frequency reuse across beams maximizes spectral efficiency. In addition, based on the centralized training and decentralized execution (CTDE) paradigm, we propose a cooperative multi-agent soft actor–critic (SAC) framework with parameter sharing and centralized training, called CoopMASAC-PSCT. Among them, an SAC agent is deployed for each beam; Specifically, all actor networks share the same architecture and parameters, and the execution phase remains decentralized, with each actor making decisions solely on its local observations; To prevent inter-cluster interference, the shared global reward integrates system throughput, queueing delay fairness, and an interference-penalty term; Moreover, only a single global critic network is employed, which accesses the joint observations and actions of all agents during training, thus balancing individual beam performance with influence between all beams. Simulation and comparative analyses demonstrate that CCBH delivers better performance while significantly reducing inference complexity.

By decoupling substrate resources, network virtualization (NV) is a promising solution for meeting diverse demands and ensuring differentiated Quality of Service (QoS). In particular, virtual network embedding (VNE) is a critical enabling technology that enhances the flexibility and scalability of network deployment by addressing the coupling of Internet processes and services. However, in the existing deep neural networks (DNNs)-based works, the closed-box nature DNNs limits the analysis, development, and improvement of systems. For example, in the Industrial Internet of Things (IIoT), there is a conflict between decision interpretability and the opacity of DNN-based methods. In recent times, interpretable deep learning (DL) represented by deep neuro fuzzy systems (DNFSs) combined with fuzzy inference has shown promising interpretability to further exploit the hidden value in the data. Motivated by this, we propose a DNFS-based VNE algorithm that aims to provide an interpretable NV scheme. Specifically, data-driven convolutional neural networks (CNNs) are used as fuzzy implication operators to compute the embedding probabilities of candidate substrate nodes through entailment operations. And, the identified fuzzy rule patterns are cached into the weights by forward computation and gradient back-propagation (BP). Moreover, the fuzzy rule base is constructed based on Mamdani-type linguistic rules using linguistic labels. In addition, the DNFS-driven five-block structure-based policy network serves as the agent for deep reinforcement learning (DRL), which optimizes VNE decision making through interaction with the environment. Finally, the effectiveness of evaluation indicators and fuzzy rules is verified by simulation experiments.

Adaptive bitrate (ABR) streaming provides an effective way to improve the Quality of Experience (QoE) of video users and is now the de facto standard for video delivery. Meanwhile, mobile edge computing (MEC) has been applied to assist ABR streaming, improving the performance of mobile networks and enabling efficient video delivery. However, smooth ABR streaming relies on the bidirectional adaptation between bitrate selection and bandwidth allocation, as they operate on distinct timescales and have different optimization goals. Moreover, since the constrained wireless resources available within a cell are shared by multiple users, their QoE should be optimized not only jointly but fairly. To this end, we propose a QoE-fairness-aware bandwidth allocation (QFA-BA) method for MEC-assisted ABR video transmission. With a novel perspective on buffer occupancy modeling, the relationship between bitrate selection and bandwidth allocation is studied. An enhanced QoE evaluation model is then proposed to correlate bitrate selection with bandwidth allocation and facilitate QFA-BA. Finally, a soft actor-critic (SAC) framework improving both the QoE and QoE-fairness is presented for QFA-BA. Compared with the state-of-the-art methods, our QFA-BA can perceive fine-grained buffer occupancy and stabilize it near a preset value with relatively more and larger bitrate switchings, exhibiting smoother convergence, better QoE (50.29%) and QoE fairness (54.81%).

With the rapid development of vehicular networks, ever-growing number of on-board sensors makes vehicle applications/tasks to be not only computation-intensive but also data-intensive. To this end, Vehicular Cloud Computing (VCC) has been convinced as a promising paradigm to offer valuable computing and sensing services to vehicles. However, considering the heterogeneity of on-board computation and sensing capabilities, how to efficiently determine the appropriate vehicle to process the task is challenging. Also, the allocation of transmission power can significantly impact the corresponding energy consumption. Therefore, to minimize the weighted sum of execution delay and energy consumption of vehicle tasks, in this paper, we propose a Collaborative Sensing-Assisted Task Offloading and Resource Al-location (CSTR) algorithm based on the Integrated Sensing and Communication (ISAC) mechanism. The optimization problem is formulated as a mixed integer nonlinear programming problem (MINLP), which is proven to be NP-hard. To achieve this, the original problem is decoupled into two sub-problems namely the task offloading problem and transmission power allocation problem, which are solved by Genetic Algorithm (GA) and convex optimization technique, respectively. Validation through several simulations based on real-world road networks has demonstrated that our proposed CSTR can outperform existing benchmark solutions under various settings.

Multimodal retrieval has received widespread consideration since it can commendably provide massive related data support for the development of computational social systems (CSSs). However, the existing works still face the following challenges: 1) rely on the tedious manual marking process when extended to CSS, which not only introduces subjective errors but also consumes abundant time and labor costs; 2) only using strongly aligned data for training, lacks concern for the adjacency information, which makes the poor robustness and semantic heterogeneity gap difficult to be effectively fit; and 3) mapping features into real-valued forms, which leads to the characteristics of high storage and low retrieval efficiency. To address these issues in turn, we have designed a multimodal retrieval framework based on web-knowledge-driven, calledunsupervised and robust graph convolutional hashing(URGCH). The specific implementations are as follows: first, a “secondary semantic self-fusion” approach is proposed, which mainly extracts semantic-rich features through pretrained neural networks, constructs the joint semantic matrix through semantic fusion, and eliminates the process of manual marking; second, a “adaptive computing” approach is designed to construct enhanced semantic graph features through the knowledge-infused of neighborhoods and uses graph convolutional networks for knowledge fusion coding, which enables URGCH to sufficiently fit the semantic modality gap while obtaining satisfactory robustness features; Third, combined with hash learning, the multimodality data are mapped into the form of binary code, which reduces storage requirements and improves retrieval efficiency. Eventually, we perform plentiful experiments on the web dataset. The results evidence that URGCH exceeds other baselines about$1\%$–$3.7\%$in mean average precisions (MAPs), displays superior performance in all the aspects, and can meaningfully provide multimodal data retrieval services to CSS.

The Industrial Internet of Things (IIoT) promotes the deep integration of new-generation communication technologies and industrial ecology. However, the popularity of computing and the proliferation of equipment scale make it a meaningful challenge to provide reasonable resource allocation for task offloading. Therefore, this article proposes a novel two-stage coordinated, distributed, and online multidomain virtual network embedding algorithm based on deep reinforcement learning (DRL)-assisted federated learning (FL) for task offloading in the IIoT. We model the IIoT as a dynamic multidomain structure and deploy local DRL servers in each factory domain combined with the distributed paradigm of FL to reduce the local resource fragmentation. Through local and global cooperation, the IIoT environment is controlled in a fine and macroscopic manner. In addition, the mechanisms of FL ensure the privacy of participant data. Finally, a comprehensive evaluation demonstrates the clear superiority of the proposed algorithm, which improves the long-term offloading revenue, resource utilization, and task offloading success rate by average 17.66%, 5.97%, and 4.52% compared to baselines, respectively.

With the maturity of 5G and Intelligent Transportation Systems (ITS) technologies and the prospect of Beyond 5G (B5G) and 6G technologies, the limited lifetime and computing of mobile devices pose significant challenges to Quality of Service (QoS). In addition, the problem of inefficient use of computing, storage, communication, and other resources still exists in communication systems. In response to the above issues, Multi-tier Computing Networks (MTCNs) migrate computationally intensive tasks to the cloud, fog, or edge with sufficient resources, thereby realizing energy-efficient collaborative computing and multi-dimensional resource sharing. However, in the MTCN environment with complex heterogeneity, and high-intensity dynamics, how to provide sustainable solutions for resource scheduling strategies is a meaningful issue. Inspired by Virtual Network Embedding (VNE) to decouple physical network configuration, we propose a multi-target-aware dynamic resource scheduling algorithm for MTCN to improve resource flexibility, which is the first attempt in this direction. Specifically, we consider differentiated QoS requirements like computing, storage, bandwidth, delay, etc., and establish multi-target-aware embedded constraints. Additionally, we present a Deep Reinforcement Learning (DRL)-based scheduling network that can interact scientifically and efficiently with the MTCN environment. It extracts environmental information as state input to better focus on dynamic characteristics as well as calculates candidate nodes and links using a three-layer network architecture and related constraints. Furthermore, the learning process is optimized through the combination of the reward mechanism and the gradient descent mechanism. Finally, comparison experiments on three widely used evaluation indicators (long-term average revenue, long-term average revenue-cost ratio, and VNR acceptance rate) verify that the proposed algorithm has made an average improvement of$19.042\%$,$2.563\%$, and$3.932\%$respectively compared with all baselines.

As electric vehicle (EV) ownership becomes more commonplace, partly due to government incentives, there is a need also to design solutions such as energy allocation strategies to more effectively support sustainable vehicle-to-grid (V2G) applications. Therefore, this work proposes an energy allocation strategy, designed to minimize the electricity cost while improving the operating revenue. Specifically, V2G is abstracted as a three-domain network architecture to facilitate flexible, intelligent, and scalable energy allocation decision-making. Furthermore, this work combines virtual network embedding (VNE) and deep reinforcement learning (DRL) algorithms, where a DRL-based agent model is proposed, to adaptively perceives environmental features and extracts the feature matrix as input. In particular, the agent consists of a four-layer architecture for node and link embedding, and jointly optimizes the decision-making through a reward mechanism and gradient back-propagation. Finally, the effectiveness of the proposed strategy is demonstrated through simulation case studies. Specifically, compared to the used benchmarks, it improves the VNR acceptance ratio, Long-term average revenue, and Long-term average revenue-cost ratio indicators by an average of 3.17%, 191.36, and 2.04%, respectively. To the best of our knowledge, this is one of the first attempts combining VNE and DRL to provide an energy allocation strategy for V2G.

Network Virtualization (NV) is an emerging network dynamic planning technique to overcome network rigidity. As its necessary challenge, Virtual Network Embedding (VNE) enhances the scalability and flexibility of the network by decoupling the resources and services of the underlying physical network. For future multi-domain physical network modeling with the characteristics of dynamics, heterogeneity, privacy, and real-time, the existing related works perform unsatisfactorily. Federated learning (FL) jointly optimizes the network by sharing parameters among multiple parties and is widely used to address data privacy and data silos. Aiming at the NV challenge of multi-domain physical networks, this work is the first to propose using FL to model VNE, and presents a VNE architecture based on Horizontal Federated Learning (HFL) (HFL-VNE). Specifically, combined with the distributed training paradigm of FL, we deploy local servers in each physical domain, which can effectively focus on local features and reduce resource fragmentation. A global server is deployed to aggregate and share training parameters, which enhances local data privacy and significantly improves learning efficiency. Furthermore, we deploy the Deep Reinforcement Learning (DRL) model in each server to dynamically adjust and optimize the resource allocation of the multi-domain physical network. In DRL-assisted FL, HFL-VNE jointly optimizes decision-making through specific local and federated reward mechanisms and loss functions. Finally, the superiority of HFL-VNE is proved by combining simulation experiments and comparing it with related works.

To realize the Interconnection of Everything (IoE) in the 6G vision, the space-based, air-based, and ground-based networks have shown a trend of integration. Compared with the traditional communications system, Space-Air-Ground Integrated Networks (SAGINs) can provide a seamless global network connection, while making full use of different network characteristics for synergy and complementarity. However, the increasing global coverage of the Internet, the growing number and variety of smart terminals, and the emergence of various high-bandwidth services have led to an explosion in communication data transmission. Despite the continuous development of communication technologies such as airborne processing and forwarding and high-throughput satellites, the quality of service (QoS) and quality of experience (QoE) for different users still cannot be guaranteed due to the power limitations of satellites and the scarcity of spectrum resources. In this work, drawing on wireless edge caching, considering that the relay of SAGIN has edge caching capability, the hot task is cached in the network nodes in advance. More, this process is optimized using distributed Deep Reinforcement Learning (DRL), thereby reducing transmission delay and relieving the pressure of task offloading on space-based networks. Compared with advanced related works, the long-term node utilization, link utilization, long-term average revenue-to-cost ratio and acceptance ratio of the proposed algorithm are increased by about 4.22%, 31.36%, 11.75% and 7.14%, respectively.

With the advantages of Millimeter wave in wireless communication network, the coverage radius and inter-site distance can be further reduced, the ultra dense network (UDN) becomes the mainstream of future networks. The main challenge faced by UDN is the serious inter-site interference, which needs to be carefully addressed by joint user association and resource allocation methods. In this paper, we propose a multi-agent Q-learning based method to jointly optimize the user association and resource allocation in UDN. The deep Q-network is applied to guarantee the convergence of the proposed method. Simulation results reveal the effectiveness of the proposed method and different performances under different simulation parameters are evaluated.