The rapid development of service computing has led to the emergence of scalable and flexible architectures such as microservices and Artificial Intelligence as a Service (AaaS), enabling the orchestration of AI-driven intelligent applications. However, existing work on intelligent applications orchestration overlooked essential microservice components that support AI services, resulting in coarse-grained and incomplete models. To ensure system integrity and enhance QoS, fine-grained collaborative orchestration of microservices and AI services is crucial. However, this poses significant challenges due to complex service dependencies, high request concurrency, and heterogeneous resource demands in edge environments. Moreover, the strong coupling between service deployment and request routing complicates their joint optimization, since effective decisions in one depend on the other. To address these challenges, we propose a collaborative orchestration framework that jointly optimizes the deployment of microservices and AI services along with probabilistic request routing in edge environments. We formulate the problem as a mixed-integer nonlinear program and leverage Jackson queuing networks for accurate delay modeling. To solve this, we develop a dual time scale hybrid greedy proximal policy optimization (DTS-HGPPO) algorithm that performs instance-level deployment and adaptive routing, enhanced with iterative instance planning, action masking and intrinsic motivation mechanisms. Extensive trace-driven experiments demonstrate that our method significantly reduces both response delay and service cost compared to state-of-the-art baselines.

Simultaneously transmitting and reflecting reconfigurable intelligent surface (STAR-RIS) can be assembled in the air-ground integrated sensing and communication (ISAC) to significantly enhance the coverage and sensing performance. However, the near-field effect should be further considered with higher carrier frequency and increasing number of STAR-RIS elements. In this paper, we propose a STAR-RIS enabled air-ground near-field ISAC scheme, where an unmanned aerial vehicle (UAV) is deployed as the mobile base station (BS) and the semi-passive STAR-RIS architecture is adopted to alleviate the severe path loss. Specifically, we maximize the weighted sum rate to guarantee both the communication and sensing functionalities by jointly modifying the beamforming vectors at the BS, the reflection/transmission matrices of the STAR-RIS and, the hovering location of the UAV to well match the near-field effect, which is non-convex with coupled variables. To address this challenge, we first decompose the problem into three subproblems via block coordinate descent. Then, the semidefinite relaxation and successive convex approximation are leveraged to recast these subproblems into convex ones. Finally, we develop an alternating algorithm with low complexity to iteratively solve them. Simulation results are shown to demonstrate the superiority and validity of the proposed scheme.

Enhancing future wireless networks presents a significant challenge for networking systems due to diverse user demands and the emergence of 6G technology. While reinforcement learning (RL) is a powerful framework, it often encounters difficulties with high-dimensional state spaces and complex environments, leading to substantial computational demands, distributed intelligence, and potentially inconsistent outcomes. Large language models (LLMs), with their extensive pretrained knowledge and advanced reasoning capabilities, offer promising tools to enhance RL in optimizing 6G wireless networks. We explore RL models augmented by LLMs, emphasizing their roles and the potential benefits of their synergy in wireless network optimization. We then examine LLM-enabled RL across various protocol layers: physical, data link, network, transport, and application layers. Additionally, we propose an LLM-assisted state representation and semantic extraction to enhance the multi-agent reinforcement learning (MARL) framework. This approach is applied to service migration and request routing, as well as topology graph generation in unmanned aerial vehicle (UAV)-satellite networks. Through case studies, we demonstrate that our framework effectively performs optimization of wireless network. Finally, we outline prospective research directions for LLM-enabled RL in wireless network optimization.

Computing Power Networks (CPNs) have become an essential network architecture for supporting emerging applications, where an efficient server deployment scheme is critical to meeting growing demands. However, existing studies on server deployment schemes overlook the significant spatiotemporal variations in renewable energy availability and electricity prices, and inadequately consider network paths and task characteristics, ultimately leading to suboptimal decisions. To bridge this gap, this paper proposes a demand-driven server deployment scheme enabled by a spatiotemporal task scheduling strategy. Firstly, for the task scheduling scheme, we leverage a demand response program to perform triple selection of computing resources, routing paths, and forwarding time. Then, for the server deployment scheme, we obtain the computing resource demand of each CPN node based on the optimized scheduling decisions and further select energy-efficient servers with low procurement costs. Due to the interdependence between the scheduling and deployment schemes, a Multi-Objective Evolutionary Algorithm (MOEA)-based hierarchical solution is developed to iteratively find the optimal deployment solution. Simulation results show that the proposed scheme significantly reduces carbon emissions and annual costs while increasing the proportion of green energy usage, outperforming benchmark methods. The findings demonstrate the effectiveness of integrating scheduling and deployment for building efficient and sustainable CPN infrastructures.

Blockchain technology is leveraged in the Internet of Things (IoT) systems to enhance data reliability and management efficiency, ensuring integrity, security, and auditability through a decentralized ledger architecture. However, resource-constrained IoT devices are unable to store the complete blockchain due to prohibitive resource consumption and performance degradation. While collaborative storage strategies have been proposed to mitigate these constraints, existing approaches often prioritize storage scalability without sufficiently addressing the selection of cooperative nodes for distributed ledger maintenance. This can lead to significant communication delays during block retrieval, undermining the real-time performance and overall efficiency of the blockchain-enabled IoT system. To address this challenge, this paper introduces a clustering-based collaborative storage scheme and proposes a novel joint optimization algorithm that iteratively refines both node clustering and block allocation strategies within the blockchain network. By structuring IoT devices into clustered peers, the algorithm reduces block query latency and facilitates efficient blockchain synchronization and update processes. Experimental evaluations confirm that the proposed method effectively alleviates storage limitations and lowers access costs in static blockchain-based IoT environments.

Generative AI (GenAI), exemplified by Large Language Models (LLMs), such as OpenAI's ChatGPT, is revolutionizing various fields. Central to this transformation is Data Center Networking (DCN), which not only provides the infrastructure support for GenAI operation, but also provisions GenAI services to users. Hence, this article explores the interplay between GenAI and DCNs, analyzing their symbiotic relationship and mutual advances. We begin by reviewing the current challenges of DCNs and GenAI-based solutions, such as data augmentation, process automation, and domain transfer. We then discuss the distinctive characteristics of GenAI workloads on DCNs, gaining insights that catalyze the evolution of DCNs to more effectively support GenAI. Moreover, to illustrate the seamless integration of GenAI with DCNs, we present a case study on GenAI-empowered DCN digital twins. Specifically, we employ an LLM equipped with retrieval augmented generation to formulate optimization problems for DCNs (e.g., resource allocation and routing) and adopt diffusion-deep reinforcement learning to solve opti-mization. The experimental results on a representative DCN optimization problem, i.e., knowledge placement, demonstrate the validity and efficiency of our proposals. We anticipate that this article can promote further research to enhance the virtuous interaction between GenAI and DCNs.

Federated learning addresses the requirements of modern communication networks, ensuring low latency, resource efficiency, and data privacy. However, challenges arise from data variability, device heterogeneity, and constrained resources. These challenges worsen with noise in the dataset and communication channel. Thus, the efficient selection of participating clients in federated learning considering all these factors is crucial. This paper proposes an efficient client selection approach integrating data quality (noise and heterogeneity), communication network quality between client and server, and cost budget considerations. Initially, we address noise in local datasets and communication channels, providing an expression to estimate the communication rounds required for convergence. Subsequently, we introduce a method to estimate the reputation of each client, considering heterogeneity in the dataset and devices. After that, we estimate the training cost and time for each client based on the calculated communication rounds and client reputation. Finally, a multi-constraint knapsack problem is formulated to select high-reputation clients within the given cost budget and time constraints. These obtained clients contribute to the federated learning process, ensuring maximum accuracy within the given constraints. The results show that with noise and heterogeneity levels surpassing 60%, the proposed approach incurs a cost at least 27% lower than existing approaches.

The rapid advancement of generative artificial intelligence (GAI) has led to the widespread adoption of AI generated content (AIGC), which generally leverages generative diffusion models (GDMs) to create high-quality multimodal content. By integrating mobile edge computing (MEC) and model compression techniques, edge AIGC can provide low latency services for mobile users (MUs) at the network edge, thereby unlocking greater potential. However, such a paradigm is vulnerable to critical security risks, such as tampering and plagiarism attacks on AIGC products. To protect the copyright of AIGC products, blockchain technology has been introduced as a promising solution. Nevertheless, most existing studies on the integrated blockchain-edge AIGC framework focus primarily on management mechanisms, while neglecting the inherent resource allocation problem for the tightly coupled AIGC inference and blockchain consensus processes, which severely limits the overall system efficiency. In this work, we propose a blockchain-enhanced edge AIGC service framework over MEC networks, where multiple edge servers (ESs) act as both model hosts generating contents for MUs and miners participating in blockchain consensus. In such a framework, we study the joint optimization of AIGC inference and blockchain consensus, including MU association, inference steps, as well as communication and computation resource allocation, aiming to maximize the total system utility. To tackle the challenges posed by mixed discrete and continuous optimization variables, we decompose the orginal problem into four sequential subproblems and propose a Joint MU association, Inference step allocation, Bandwidth allocation, Power allocation, and Streaming multiprocessor allocation (JMIBPS) algorithm based on Markov approximation, interior point method, and successive convex approximation. Simulation results validate the effectiveness of the proposed algorithm, which improves the total system utility by at least 10.21% compared with representative benchmark schemes.

The growing demand for maritime broadband services exposes the limitations of conventional satellite–ground systems, which often fail to jointly optimize spatial–temporal resources, computation costs, and the trade-offs between energy and information freshness. To address these challenges, we develop a physics-aware system model for near-shore Space-Air-Ground-Sea Integrated Networks (SAGSIN), capturing realistic maritime channels, Poisson task arrivals, Age of Information (AoI), and energy consumption. Based on this model, we propose a AI-enabled dual-timescale framework based on heterogeneous Mixture-of-Experts (MoE). Specifically, the framework operations are formulated as a two-agent Markov decision process: the long-slot agent determines satellite beam-hopping coverage, while the short-slot agent jointly optimizes per-slot task offloading, UAV deployment, and resource allocation. Both agents employ MoE policies with Top-K gating. Simulation results validate that the proposed framework achieves superior AoI–energy trade-offs compared with baseline methods, highlighting the potential of AI-driven autonomy in next-generation network orchestration.

Mixture of Experts (MoE) has emerged as a promising paradigm for scaling model capacity while preserving computational efficiency, particularly in large-scale machine learning architectures such as large language models (LLMs). Recent advances in MoE have facilitated its adoption in wireless networks to address the increasing complexity and heterogeneity of modern communication systems. This paper presents a comprehensive survey of the MoE framework in wireless networks, highlighting its potential in optimizing resource efficiency, improving scalability, and enhancing adaptability across diverse network tasks. We first introduce the fundamental concepts of MoE, including various gating mechanisms and the integration with generative AI (GenAI) and reinforcement learning (RL). Subsequently, we discuss the extensive applications of MoE across critical wireless communication scenarios, such as vehicular networks, unmanned aerial vehicles (UAVs), satellite communications, heterogeneous networks, integrated sensing and communication (ISAC), and mobile edge networks. Furthermore, key applications in channel prediction, physical layer signal processing, radio resource management, network optimization, and security are thoroughly examined. Additionally, we present a detailed overview of open-source datasets that are widely used in MoE-based models to support diverse machine learning tasks. Finally, this survey identifies crucial future research directions for MoE, emphasizing the importance of advanced training techniques, resource-aware gating strategies, and deeper integration with emerging 6G technologies.

Unmanned aerial vehicles (UAVs) are increasingly used in smart city communications for air-ground communications due to their flexibility, low cost, and independence from ground conditions, enabling high data rates for future networks. This paper explores UAV-to-vehicle (U2V) mmWave integrated sensing and communication (ISAC), where vehicles are represented as rigid shapes in a 3D radar point cloud. Considering maximizing channel capacity with multi-user interference and radar performance, two adaptive optimization problems are proposed, incorporating vehicle-to-vehicle (V2V) communication for interference mitigation. The radar point cloud-driven reinforcement learning (PointRL) algorithm is designed to solve these problems. It includes a point cloud-based deep neural network (PDNN) for extracting action spaces from 3D radar data and a decision network that reduces network complexity through segmentation and connection. A linear weighted sliding window reward mechanism is also designed to enhance decision-making in dynamic environments. Simulation results show that the proposed PointRL outperforms benchmark methods.

AI-enabled wireless communications have attracted tremendous research interest in recent years, particularly with the rise of novel paradigms such as low-altitude integrated sensing and communication (ISAC) networks. Within these systems, feature engineering plays a pivotal role by transforming raw wireless data into structured representations suitable for AI models. Hence, this article offers a comprehensive investigation of feature engineering techniques in AI-driven wireless communications. Specifically, we begin with a detailed analysis of fundamental principles and methodologies of feature engineering. Next, we present its applications in wireless communication systems, with special emphasis on ISAC networks. Finally, we introduce a generative AI-based framework, which can reconstruct signal feature spectrum under malicious attacks in low-altitude ISAC networks. The case study shows that it can effectively reconstruct the signal spectrum, achieving an average structural similarity index improvement of 4%, thereby supporting downstream sensing and communication applications.

Dear Editor, This letter addresses the critical challenge of preserving privacy in graph learning without compromising on data utility.Differential pri-vacy(DP)is emerging as an effective method for privacy-preserving graph learning.However,its application often diminishes data utility,especially for nodes with fewer neighbors in graph neural networks(GNNs).Given that most real-world graph data follow a power-law distribution with a majority of low-degree nodes,we propose person-alized differential privacy graph neural network(PDPGNN),a novel GNN training method.The novelty of PDPGNN lies in uniquely offering personalized differential privacy by allocating privacy bud-gets based on node degrees,effectively improving the data utility for nodes with fewer connections.Additionally,PDPGNN integrates a weighted aggregation mechanism to enhance model accuracy.Theo-retical analysis shows that PDPGNN can achieve ∈-differential pri-vacy for graph data,making a balance between privacy protection and data utility.Experimental results on four real-world graph datasets demonstrate the effectiveness of PDPGNN.

Generative artificial intelligence (GAI) has achieved remarkable progress across diverse scenarios by combining generalized intelligence derived from large-scale web datasets with specialized expertise obtained from domain-specific data. However, current patterns of data collection and utilization suffer from inconsistent quality, infrequent updates, data silos, and limited sharing, which constrain the sustainable development of GAI. In this context, this article proposes a next-generation generative data infrastructure (NG2DI) framework from a data-centric perspective to achieve the synergistic co-evolution of data ecosystems and GAI. Specifically, we first introduce the layered architecture of NG2DI and explain how it supports each stage of the data lifecycle. We then analyze the data flow dynamics driven by interactions among NG2DI participants and highlight the related technical implementation tutorial. Furthermore, to illustrate the bidirectional optimization between data ecosystems and GAI, we demonstrate that NG2DI can facilitate forward optimization of GAI capabilities and enable reverse feedback from optimized GAI to enhance NG2DI functions. To validate these dual capabilities, we present two case studies in industrial Internet of Things (IIoT) scenarios: one demonstrates NG2DI-enabled data trading and GAI agent fine-tuning, while the other evaluates GAI-enabled data synthesis and scheduling strategy generation for NG2DI.

As edge applications demand real-time processing with limited bandwidth and energy, traditional communication systems face challenges to meet performance requirements due to the centralized architecture and redundant data transmission. To address these challenges, we propose a UAV-assisted semantic edge computing network that leverages UAV mobility and semantic communication. We formulate a joint optimization problem involving UAV trajectory, data allocation, and semantic extraction to maximize the semantic processing rate. To solve this problem, we develop a hybrid deep deterministic policy gradient (H-DDPG) algorithm that integrates deep reinforcement learning (DRL) with convex optimization via block coordinate descent (BCD), thereby enabling efficient joint decision-making across tightly coupled variables. Furthermore, we propose a hybrid diffusion deep deterministic policy gradient (H-D3PG) algorithm, which incorporates denoising diffusion models into the DRL framework. By addressing the limited adaptability of deterministic strategies, this design enhances policy expressiveness and stability. As a result, the algorithm enables adaptive trajectory control under time-varying semantic tasks and wireless channel conditions in UAV-assisted edge networks. Simulations show that H-D3PG improves the semantic processing rate by up to 38.8% while reducing energy consumption compared to Raw Data Transmission.

With the increasing emphasis on data privacy, federated learning (FL) networks show great potential through distributed training without directly sharing the raw data. However, the coverage of FL terrestrial servers is usually limited. Therefore, we propose leveraging the unmanned aerial vehicle (UAV) as a mobile flying server to further improve the wireless coverage and training efficiency of FL. Nevertheless, frequent exchanges of model parameters in UAV-assisted FL can result in serious security risks. In order to achieve this, we propose a covert FL scheme assisted by the UAV, which can protect the transmission key features of local models from being detected by wardens. Furthermore, we discuss the FL network and covert communications, as well as the exceptional features of UAV-assisted covert FL. Finally, we present two case studies, one of which focuses on minimising the network latency, while the other concentrates on reducing the energy consumption of covert FL. The simulation results demonstrate the efficacy of the proposed schemes in addressing future challenges.

Digital twin-enabled vehicular edge computing leverages virtual representations of physical vehicles to facilitate real-time intelligent services such as collaborative perception, trajectory prediction, fault diagnosis, and proactive decision-making through seamless integration of sensing, computing, and communication resources. The effectiveness of these services critically depends on the timely execution of deep learning inference tasks. However, vehicular mobility and dynamic contention for limited edge resources lead to significant delays in inference execution, thereby degrading the accuracy of digital twin-assisted decision-making. To address this challenge, we propose a hierarchical framework for collaborative inference and digital twin deployment. At the upper layer, a mobility-aware strategy for inference task offloading and digital twin deployment is employed, which incorporates dynamic action-space pruning to select edge nodes based on vehicular mobility patterns and link stability. At the lower layer, a delay-resource gradient-based optimization mechanism is designed to adaptively allocate computational resources by matching dynamic inference demands with available edge capacity. Extensive simulations demonstrate the effectiveness of the proposed approach, showing a reduction of at least 15% in Age of Information (AoI) compared to the baseline methods. Moreover, our approach significantly improves the inference efficiency and system resource utilization, and reduces the deadline violation rate.

The rapid development of AI accelerates the implementation and delivery of AI applications in diverse fields. In cloud-edge collaboration, delivering a complete AI application relies on the robust coordination between AI-supporting microservices and AI services. However, most existing studies only coarsely considered monolithic AI service orchestration while neglecting microservice orchestration. Such coarse-grained orchestration severely impacts application performance. To enable diverse high-performance AI applications, fine-grained hybrid orchestration of AI services and microservices (HOAIM) is highly desirable, yet presents formidable challenges. Due to heterogeneous services, call dependencies, and service multiplexing, fine-grained hybrid orchestration modeling is highly non-trivial. Moreover, the tight coupling between deployment and routing results in a complex joint optimization problem. To address this, we first propose a heterogeneous service orchestration network that supports orchestration optimization and automated management. Then, based on queuing networks and multi-instance models, we conduct an accurate analysis of delay and load. Furthermore, to achieve efficient hybrid orchestration, we propose preference-driven resource allocation and instance computation algorithms, along with reinforcement learning with action masking and reward shaping. Finally, extensive trace-driven simulations demonstrate that our algorithms optimize average response delay by up to 41.83%, and achieve significant advantages in load balancing, response success rate, and resource efficiency.

Orthogonal time frequency space (OTFS) modulation offers strong resilience to Doppler effects but suffers from high system latency, limiting its use in low-latency communications. This paper proposes a channel estimation algorithm for low-latency OTFS systems with large delay spreads and fractional Doppler effects. In the delay-time (DT) domain, impulse pilots are placed at equidistant intervals along the first row of the DT grid to eliminate interference from aliased delays, and a threshold detection method estimates the delays. Doppler shifts and path gains are then estimated using the discrete Fourier transform (DFT). Simulation results show that the proposed algorithm achieves near-optimal bit error rate (BER) under sparse delay spreads.