In this paper, we propose a novel mobility-aware resource provisioning scheme for edge-assisted extended reality (XR) services. The goal is to minimize resource consumption while satisfying user quality of experience (QoE) requirement, which is measured by the weighted sum of visual quality, quality variation, and round-trip interaction latency. Specifically, we present a mobility model to capture both user spatial movements and XR content interaction features. Since user viewing distance and interaction time are key model parameters that affect the spatiotemporal service demand for XR content rendering and delivery at the edge, we estimate user-specific model parameters and adopt a sample average approximation method to model the relationship between user QoE and the consumption of both communication and edge computing resources. We design a coordinate descent algorithm to make resource provisioning decisions, where a deep neural network provides a valuable initial point to accelerate convergence. Simulation results demonstrate that our proposed scheme is more efficient to utilize network resources in comparison with benchmark schemes while satisfying user QoE requirements.

In this paper, we propose a novelpolarforming antenna (PA)to achieve cost-effective wireless sensing and communication. Specifically, the PA can enable polarforming to adaptively control the antenna’s polarization electrically as well as tune its position/rotation mechanically, so as to effectively exploit polarization and spatial diversity to reconfigure wireless channels for improving sensing and communication performance. To analyze the performance gain of PA, we study a PA-enhanced integrated sensing and communication (ISAC) system that utilizes user location sensing to facilitate communication between a PA-equipped base station (BS) and PA-equipped users, by focusing on a new practical channel setup where the locations of users are nearly time-invariant but their orientations may change frequently (e.g., mobile phones rotated by spectators seated in a stadium while taking live photos). First, we model the PA channel in terms of transceiver antenna polarforming vectors and antenna positions/rotations. We then propose a two-timescale ISAC protocol, where in the slow timescale, user localization is first performed, followed by the optimization of the BS antennas’ positions and rotations based on the sensed user locations; subsequently, in the fast timescale, transceiver polarforming is adapted to cater to the instantaneous orientation of user devices in three-dimensional (3D) space, with the optimized BS antennas’ positions and rotations. We propose a new polarforming-based user localization method that uses a structured time-domain pattern of pilot-polarforming vectors to extract the common stable components in the PA channel across different polarizations based on the parallel factor (PARAFAC) tensor model. Moreover, we maximize the achievable average sum-rate of users by jointly optimizing the fast-timescale transceiver polarforming, including phase shifts and amplitude variations, along with the slow-timescale antenna rotations and positions at the BS. Simulation results validate the effectiveness of polarforming-based localization algorithm and demonstrate the performance advantages of polarforming, antenna placement, and their joint design in comparison with various benchmarks without polarforming or antenna position/rotation adaptation.

The Internet of Vehicles (IoV) is undergoing a transformative evolution, enabled by advancements in future 6 G network technologies, to support intelligent, highly reliable, and low-latency vehicular services. However, the enhanced capabilities of loV have heightened the demands for efficient network resource allocation while simultaneously giving rise to diverse vehicular service requirements. For network service providers (NSPs), meeting the customized resource-slicing requirements of vehicle service providers (VSPs) while maximizing social welfare has become a significant challenge. This paper proposes an innovative solution by integrating a mean-field multi-agent reinforcement learning (MFMARL) framework with an enhanced Vickrey-Clarke-Groves (VCG) auction mechanism to address the problem of social welfare maximization under the condition of unknown VSP utility functions. The core of this solution is introducing the “value of information” as a novel monetary metric to estimate the expected benefits of VSPs, thereby ensuring the effective execution of the VCG auction mechanism. MFMARL is employed to optimize resource allocation for social welfare maximization while adapting to the intelligent and dynamic requirements of IoV. The proposed enhanced VCG auction mechanism not only protects the privacy of VSPs but also reduces the likelihood of collusion among VSPs, and it is theoretically proven to be dominant-strategy incentive compatible (DSIC). The simulation results demonstrate that, compared to the VCG mechanism implemented using quantization methods, the proposed mechanism exhibits significant advantages in convergence speed, social welfare maximization, and resistance to collusion, providing new insights into resource allocation in intelligent 6 G networks.

Future 6G networks are envisioned to facilitate edge-assisted mobile augmented reality (MAR) via strengthening the collaboration between MAR devices and edge servers. In order to provide immersive user experiences, MAR devices must timely upload camera frames to an edge server for simultaneous localization and mapping (SLAM)-based device pose tracking. In this paper, to cope with user-specific and non-stationary uplink data traffic, we develop a digital twin (DT)-based approach for user-centric communication service provision for MAR. Specifically, to establish DTs for individual MAR devices, we first construct a data model customized for MAR that captures the intricate impact of the SLAM-based frame uploading mechanism on the user-specific data traffic pattern. We then define two DT operation functions that cooperatively enable adaptive switching between different data-driven models for capturing non-stationary data traffic. Leveraging the user-oriented data management introduced by DTs, we propose an algorithm for network resource management that ensures the timeliness of frame uploading and the robustness against inherent inaccuracies in data traffic modeling for individual MAR devices. Trace-driven simulation results demonstrate that the user-centric communication service provision achieves a 14.2% increase in meeting the camera frame uploading delay requirement in comparison with the slicing-based communication service provision widely used for 5G.

In this paper, we propose a novel service continuityaware Service Function Chain (SFC) embedding scheme for dynamic large-scale LEO satellite networks, where service disruptions occur when satellites hosting virtual network functions of an SFC move out of the service region. Particularly, we define a new metric, i.e., the Remaining Time to Migration (RTTM), which indicates the remaining functional time of an SFC before SFC reconfiguration is needed. We then formulate a service continuity-aware SFC embedding problem with the objective of maximizing the long-term acceptance ratio while increasing the normalized RTTM of accepted SFCs. We propose a scalable graph neural network-assisted deep reinforcement learning (DRL) approach to solve the embedding problem. By employing a differentiable pooling technique, we condense the feature representation of large-scale LEO satellite networks, thereby reducing the computational complexity of the down-stream DRL-based decision-making. Simulation results show that our approach reduces the proportion of reconfigured SFCs by 60 % compared to the benchmark, indicating its effectiveness in enhancing service continuity.

This paper presents a novel model-assisted learning scheme for resource allocation in environment-aware mobile augmented reality (AR) content delivery. The goal is to minimize the long-term communication resource consumption for delivering virtual content visible to an individual AR user by optimizing the communication resource allocation for user positioning and environment mapping. In specific, we first develop a mathematical model to estimate the content visibility uncertainty and the content delivery resource consumption. We then generate a reference resource allocation decision that guides a deep reinforcement learning-based decision process to efficiently adapt to non-stationary user and environment dynamics. We conduct trace-driven simulations to evaluate the performance of the proposed scheme, and the results demonstrate that, the proposed scheme significantly reduces communication resource consumption for delivering virtual content visible to an individual AR user, compared to benchmark schemes.

To fulfill future diverse user requirements, 6G networks are envisioned to provide everyone-centric customized services ubiquitously and precisely. However, the diversity in user requirements and the heterogeneity in network resources challenge conventional network operators in network management and service provision. In this article, we investigate the artificial intelligence (AI) service provision in the multilayer heterogeneous network. To provide ubiquitous intelligence to users with different computing requirements, an intelligence-native network architecture is designed. Based on the proposed architecture and the AI model stitching mechanism, we formulate the joint AI provision and access selection problem as a mixed integer nonlinear programming (MINLP) problem to maximize the average user satisfaction value and user satisfaction rate. Then, a heuristic solution based on Dung Beetle algorithm is proposed to optimize the AI model selection, AI service deployment, user access, and stitching coefficient jointly. Extensive simulations are conducted to evaluate the performance of our proposed architecture and algorithm.

Presents corrections to the paper, (Correction to “Toward Native Intelligence: An Efficient and Flexible AI Services Provision Scheme in Multilayer Heterogeneous Networks”).

In digital-twin-enabled heterogeneous vehicular networks (DT-HetVNets), vehicles need to synchronize data to their DTs deployed in the cloud for decision-making. However, for a vehicle which is simultaneously covered by a group of heterogeneous network infrastructures, the DT of the vehicle (DT-V) can connect with the DTs of infrastructures (DT-Is) in different infrastructure groups across regions in the virtual networks so that each DT-V may deceive the DT-Is by interacting with multiple DT-I groups and selecting the optimal one to synchronize data. To this end, we propose an on-demand data synchronization scheme for DT-Vs and DT-Is. In the scheme, infrastructures and vehicles are grouped based on their geographical locations and the arrival time of each vehicle through which the DT-Vs and DT-Is can interact with each other to make decisions in groups. Then, the requirements of DT-Vs (i.e., minimize synchronization cost and maximize synchronization satisfaction) and DT-Is (i.e., maximize profits) are considered to design their utility functions and the decision-making process between the DT-Vs in each group and the DT-Is in each group is formulated as a Stackelberg game to obtain their optimal strategies. After that, considering the deceptive behavior of vehicles, a joint optimization algorithm that integrates the Stackelberg game and the selection of each DT-V is designed to obtain the real equilibrium solution for DT-Vs and DT-Is to maximize their utilities. Simulation results show that our scheme can obtain the highest utilities compared with the traditional schemes.

Continual machine learning in the context of limited computational resources and data availability is critical in the connected digital world. Current intelligent applications predominantly rely on deep learning models requiring labor/computation-intensive training. These models often struggle to adapt effectively to new data while preserving performance on previously learned knowledge. In this paper, we introduce a lightweight method for continual knowledge adaptation that can address these challenges. To prevent disruption of the existing services, we propose a Mixture-of-Experts (MoE) adapter that integrates seamlessly with the existing vision model to encode new data. The weights of the original model are kept fixed during the adaptation process, ensuring the preservation of previously learned knowledge. The MoE technique enables scaling up the parameters of the adapter while maintaining a relatively low computation, making it fit for constrained devices in mobile computation scenarios. Furthermore, to enhance learning efficiency and accelerate convergence with new data, we implement a knowledge fusion mechanism that facilitates interaction between the existing knowledge and the information extracted from new data. The timing of employing the fusion module is further investigated. We find that it is conducive in scenarios where the task's performance requirements are enhanced. The MoE adapter and knowledge fusion module are integrated at each stage with minimal trainable parameters, efficiently optimizing resource usage. Extensive experiments and ablation studies validate the effectiveness of the proposed method. Specifically, the proposed method prevents an accuracy drop of 43.02% on the previous data compared to the continual train method, while achieving an accuracy of 44.81% on the new data, which is even 0.34% higher than fully training a new model.

Cellular operators face significant challenges in cutting operating expenses while maintaining the quality of service (QoS) for users due to growing network traffic and dynamic user connections. These challenges are addressed by the cloud radio access network (C-RAN) architecture, which includes a centralized pool of baseband units (BBUs) and distributes them from remote radio heads (RRHs). The key to improving C-RAN performance is to dynamically allocate large-scale RRHs to different BBUs in real time. In this paper, we propose a user behavior-aware RRH-BBU mapping framework to improve the performance of large-scale C-RANs by predicting RRH traffic and users in advance. First, we propose a Multivariate RRH time series Prediction Model (MRPM) that captures the spatio-temporal patterns in the data to predict the traffic volume and the number of users of RRHs, which represents key indicators of RRH connection states. Second, we formulate the RRH-BBU mapping as a Markov decision process problem to optimize cost and QoS by considering BBU utilization, BBU energy consumption, RRH migration frequency, and BBU load balancing. Third, we propose a prediction-based RRH-BBU mapping scheme (PB-RBM) to find the optimal RRH-BBU mapping strategy by leveraging the prediction information of MRPM. In the PB-RBM algorithm, we employ an A3C algorithm to learn the mapping policy and group the RRHs based on a defined popularity metric to reduce the state and action space of the reinforcement learning algorithm. Finally, extensive experiments are conducted on a real-world dataset, and our algorithm is compared with several matching algorithms, such as ACKTR, heuristic, etc., to demonstrate its superiority, especially reducing 17.5% in RMSE compared to the best-performing baseline.

The demand for continual machine learning in the context of limited computational resources and data availability is critical in the evolving landscape of the connected digital world. Current network applications predominantly rely on deep learning models that require labor/computation-intensive training processes. These models often struggle to effectively adapt to new data while preserving performance on previously acquired knowledge. In this paper, we introduce a lightweight framework for continual knowledge adaptation and learning designed to address these challenges. To prevent disruption of existing services, we propose an attention-based adapter that integrates seamlessly with the existing vision model to encode new incoming data. The weights of the original model are kept fixed during the adaptation process, ensuring the preservation of previously learned knowledge. Furthermore, to enhance learning efficiency and accelerate convergence with new data, we implement a knowledge fusion mechanism that facilitates interaction between existing knowledge and information from new data. Our framework is modular, enabling flexible deployment across distributed devices. The adapter and knowledge fusion module are implemented at each stage with minimal trainable parameters, optimizing resource usage. Extensive experiments and ablation studies validate the effectiveness of the proposed framework.

Resource slicing in low Earth orbit satellite networks (LSN) is essential to support diversified services. In this paper, we investigate a resource slicing problem in LSN to reserve resources in satellites to achieve efficient resource provisioning. To address the challenges of non-stationary service demands, inaccurate prediction, and satellite mobility, we propose an adaptive digital twin (DT)-assisted resource slicing scheme for robust and adaptive resource management in LSN. Specifically, a slice DT, being able to capture the service demand prediction uncertainty through collected service demand data, is constructed to enhance the robustness of resource slicing decisions for dynamic service demands. In addition, the constructed DT can emulate resource slicing decisions for evaluating their performance, enabling adaptive slicing decision updates to efficiently reserve resources in LSN. Simulation results demonstrate that the proposed scheme outperforms benchmark methods, achieving low service demand violations with efficient resource consumption.

Satellite-terrestrial integrated networks (STIN) are envisioned as a promising architecture for ubiquitous network connections to support diversified services. In this paper, we pro-pose a novel resource slicing scheme with cross-cell coordination in STIN to satisfy distinct service delay requirements and efficient resource usage. To address the challenges posed by spatiotemporal dynamics in service demands and satellite mobility, we formulate the resource slicing problem into a long-term optimization problem and propose a distributed resource slicing (DRS) scheme for scalable and flexible resource management across different cells. Specifically, a hybrid data-model co-driven approach is developed, including an asynchronous multi-agent reinforcement learning- based algorithm to determine the optimal satellite set serving each cell and a distributed optimization-based algorithm to make the resource reservation decisions for each slice. Simulation results demonstrate that the proposed scheme outperforms benchmark methods in terms of resource usage and delay performance.

In this paper, we propose a digital twin (DT)-based user-centric approach for processing sensing data in an integrated sensing and communication (ISAC) system. The considered scenario involves an ISAC device with a lightweight deep neural network (DNN) and a mobile edge computing (MEC) server with a large DNN. After collecting sensing data, the ISAC device either processes the data locally or uploads them to the server for higher-accuracy data processing. To cope with data drifts, the server updates the lightweight DNN when necessary, referred to as continual learning. Our objective is to minimize the long-term average computation cost of the MEC server by jointly optimizing two decisions, i.e., sensing data offloading and sensing data selection for the DNN update. A DT of the ISAC device is constructed to predict the impact of potential decisions on the long-term computation cost of the server, based on which the decisions are made with closed-form formulas. Experiments on executing DNN-based human motion recognition tasks are conducted to demonstrate the outstanding performance of the proposed DT-based approach in computation cost minimization.

This paper introduces a real-time collaborative sensing scheme for wireless sensor networks in time-varying environments. The objective is to maximize the sensors' performance by effectively allocating communication resources for data sharing. Specifically, we utilize digital twins (DTs) to characterize dynamic collaborative sensing demands for each sensor through data-driven methods. Building on the DT design, we propose a resource allocation scheme to optimize the communication resources allocated at each stage of collaborative sensing and determine the most effective collaborative sensing policy. By profiling sensors using DTs, the network controller can effectively coordinate the sensors without exhaustively exploring all collaborative sensing policies. Numerical results demonstrate the effectiveness of our proposed scheme in optimizing the sensing performance for all sensors.

As an integral component of the space-air-ground integrated network (SAGIN), the low Earth orbit (LEO) satellite network has displayed immense potential in providing ubiquitous connectivity and broadband mobile communication. However, the intrinsic dynamics of LEO satellites pose unprecedented challenges in network management and service delivery. In this paper, we investigate the service function chain (SFC) orchestration in dynamic LEO satellite networks to achieve flexible and efficient service provision. Considering the service requirements and the limitations of network resources, we formulate the SFC orchestration problem as the integer nonlinear programming (INLP) problem for maximizing the service acceptance and the load fairness of satellites. Then, an efficient heuristic algorithm is proposed to solve this problem. Addressing the situation with frequent service requests, a graph attention network (GAT)-based approach with low complexity is also presented. Simulation results demonstrate that our proposed approaches outperform the benchmarks by a substantial margin in terms of load fairness and service acceptance. Besides, the proposed GAT-based approach shows its advantage in computation complexity, and exhibits robustness in unstable network scenarios with intermittent link interruptions.

Device-edge collaboration on deep neural network (DNN) inference is a promising approach to efficiently utilizing network resources for supporting Artificial Intelligence of Things (AIoT) applications. In this article, we propose a novel digital twin (DT)-assisted approach to device-edge collaboration on DNN inference that determines whether and when to stop local inference at a device and upload the intermediate results to complete the inference on an edge server. Instead of determining the collaboration for each DNN inference task only upon its generation, multi-step decision making is performed during the on-device inference to adapt to the dynamic computing workload status at the device and the edge server. To enhance the adaptivity, a DT is constructed to evaluate all potential offloading decisions for each DNN inference task, which provides augmented training data for a machine learning-assisted decision-making algorithm. Then, another DT is constructed to estimate the inference status at the device to avoid frequently fetching the status information from the device, thus reducing the signaling overhead. We also derive necessary conditions for optimal offloading decisions to reduce the offloading decision space. Simulation results demonstrate the outstanding performance of our DT-assisted approach in terms of balancing the tradeoff among inference accuracy, delay, and energy consumption.

In this paper, we propose a joint sensing and communication framework for virtual reality (VR) applications in Metaverse. Although millimeter-wave (mmWave) communication can achieve multi-Gbps wireless transmission data rate, a slight movement of the VR headset can result in a significant drop in transmission rate. This significantly deteriorates user’s experience in Metaverse applications. By characterizing the relationship between mmWave beam gain and beam width, we find that adaptively turning off part of antennas can improve the overall transmission performance for mobile Metaverse. To this end, we formulate a problem with the objective of adaptively configuring receiver’s phase shift, and adjusting the beam width to cope with the variation in VR user’s viewpoints in Metaverse services. By revealing the correlation between power consumption and signal-to-noise ratio of VR headset, the proposed dual method based on meta reinforcement learning enables reliable and energy-efficient mmWave communication for VR. Based on the sensing information collected from VR users, the beamforming strategy is continuously updated by reshaping the reward of learning process, which minimizes the power consumption while meeting transmission requirements of Metaverse applications. Extensive experimental results demonstrate that the adaptability of the proposed framework outperforms the existing benchmarks in various VR scenarios, which ensures the applicability of mmWave communication to Metaverse applications.

This article introduces an on-demand collaborative sensing scheme for industrial Internet of Things (IIoT) sensors in time-varying sensing environments, aiming to optimize the sensing performance by effectively allocating communication resources for sensory data sharing. Particularly, we propose a novel digital twins (DTs)-empowered resource allocation solution to facilitate scalable and flexible collaborative sensing. First, DTs create mathematical models using real-time network data to characterize the dynamic resource demands in collaborative sensing. Second, the performance of mathematical models in DTs is evaluated through data-driven methods. Building on our DT design, we propose a joint collaborative sensing and DT management scheme to optimize the resource allocation for sensory data sharing and DT operation. Furthermore, we develop a DT evaluation method featuring a variational autoencoder to evaluate the accuracy of DTs and enable closed-loop DT-based resource allocation. Numerical results demonstrate the effectiveness of our proposed collaborative sensing scheme in optimizing the sensing performance for all sensors.

Edge-device collaboration has the potential to facilitate compute-intensive device pose tracking for resource-constrained mobile augmented reality (MAR) devices. In this article, we devise a 3-D map management scheme for edge-assisted MAR, wherein an edge server constructs and updates a 3-D map of the physical environment by using the camera frames uploaded from an MAR device, to support local device pose tracking. Our objective is to minimize the uncertainty of device pose tracking by periodically selecting a proper set of uploaded camera frames and updating the 3-D map. To cope with the dynamics of the uplink data rate and the user’s pose, we formulate a Bayes-adaptive Markov decision process problem and propose a digital twin (DT)-based approach to solve the problem. First, a DT is designed as a data model to capture the time-varying uplink data rate, thereby supporting 3-D map management. Second, utilizing extensive generated data provided by the DT, a model-based reinforcement learning algorithm is developed to manage the 3-D map while adapting to these dynamics. Numerical results demonstrate that the designed DT outperforms Markov models in accurately capturing the time-varying uplink data rate, and our devised DT-based 3-D map management scheme surpasses benchmark schemes in reducing device pose tracking uncertainty.

In distributed IoT data systems, full-size data collection is impractical due to the energy constraints and large system scales. Our previous work has investigated the advantages of integrating matrix sampling and inference for compact distributed IoT data collection, to minimize the data collection cost while guaranteeing the data benefits. This paper further advances the technology by boosting fast and accurate inference for those distributed IoT data systems that are sensitive to computation time, training stability, and inference accuracy. Particularly, we proposeCODE$^{+}$+, i.e.,Compact Distributed IOTData CollEction Plus, which features a cluster-based sampling module and a Convolutional Neural Network (CNN)-Transformer Autoencoders-based inference module, to reduce cost and guarantee the data benefits. The sampling component employs a cluster-based matrix sampling approach, in which data clustering is first conducted and then a two-step sampling is performed in accordance with the number of clusters and clustering errors. The inference component integrates a CNN-Transformer Autoencoders-based matrix inference model to estimate the full-size spatio-temporal data matrix, which consists of a CNN-Transformer encoder that extracts the underlying features from the sampled data matrix and a lightweight decoder that maps the learned latent features back to the original full-size data matrix. We implementCODE$^{+}$+under three operational large-scale IoT systems and one synthetic Gaussian distribution dataset, and extensive experiments are provided to demonstrate its efficiency and robustness. With a 20% sampling ratio,CODE$^{+}$+achieves an average data reconstruction accuracy of 94% across four datasets, outperforming our previous version of 87% and state-of-the-art baseline of 71%.

Federated Learning (FL) is a promising paradigm widely used in privacy-preserving machine learning. It enables distributed devices to collaboratively train a model while avoiding data transfer between clients. Nevertheless, FL suffers from bottlenecks in training speed due to client heterogeneity, resulting in increased training latency and server aggregation lagging. To address this issue, a novel Split Federated Learning (SFL) framework is proposed. It pairs clients with different computational resources based on their computational resources and inter-client communication rates. The neural network model is split into two parts at the logical level, and each client computes only its assigned part using Split Learning (SL) to accomplish forward inference and backward training. Besides, a heuristic greedy algorithm is proposed to effectively deal with the client pairing problem by reconstructing the training latency optimization as a graph edge selection problem. Simulation results show that the proposed method can significantly improve the FL training speed and achieve high performance in both independent identical distribution (IID) and Non-IID data distribution.

In this paper, we propose a digital twin (DT)-assisted cloud-edge collaborative transcoding scheme to enhance user satisfaction in live streaming. We first present a DT-assisted transcoding workload estimation (TWE) model for the cloud-edge collaborative transcoding. Particularly, two DTs are constructed for emulating the cloud-edge collaborative transcoding process by analyzing spatial-temporal information of individual videos and transcoding configurations of transcoding queues, respectively. Two light-weight Bayesian neural networks are adopted to fit the TWE models in DTs, respectively. Moreover, we formulate a transcoding-path selection problem to maximize long-term user satisfaction within an average service delay threshold, taking the dynamics of video arrivals and video requests into account. The problem is transformed into a standard Markov decision process by using the Lyapunov optimization, which is further solved by a deep reinforcement learning algorithm. Simulation results based on the real-world dataset demonstrate that the proposed scheme can effectively enhance user satisfaction compared with benchmark schemes.

In this paper, we investigate joint caching and computing resource reservation for supporting location-aware augmented reality (AR) applications in an edge-assisted two-tier radio access network. We aim at minimizing the caching and computing resource consumption while satisfying the AR service delay requirement. Specifically, to capture the spatio-temporal AR service dynamics, the resource consumption minimization problem is formulated as a long-term stochastic optimization problem. Due to the time-varying service demands and tightly coupled multi-resource reservation decisions, we propose a novel resource reservation algorithm based on the Lyapunov optimization technique to solve the problem. We first transform the original long-term problem into multiple one-shot optimization problems, each of which is then solved by our designed iterative algorithm in an online manner. Simulation results demonstrate that the proposed algorithm can significantly reduce the overall resource consumption compared to benchmark algorithms.

As an integral component of space-air-ground integrated networks (SAGINs), the low Earth orbit (LEO) satellite networks have displayed immense potential in providing ubiquitous connectivity and broadband mobile communication. However, the intrinsic dynamics of LEO satellites poses unprecedented challenges in network management, multi-dimensional resource scheduling, and service delivery. In this paper, we study the service function chain (SFC) orchestration in dynamic LEO satellite networks, with the aim of achieving flexible and efficient service provision. Considering the service requirements and the load fairness of LEO satellite networks, we formulate the SFC deployment problem as an integer nonlinear programming (INLP) problem. We then introduce a load-aware SFC orchestration algorithm to improve serving capacity and load fairness. Additionally, we address the issue of SFC migration in dynamic LEO satellite networks to ensure service continuity. To minimize the service interruption and network resource wastes, a Tabu search (TS)-based approach is presented to optimize the virtual network function (VNF) migration. Simulation results demonstrate that our proposed approaches outperform the benchmark by a substantial margin in terms of load fairness, without compromising service acceptance.

In this paper, we propose collaborative beamforming (CB) in unmanned aerial vehicle (UAV)-assisted communication networks to improve transmission data rate with minimum energy consumption. Specifically, CB allows a group of UAVs forming a virtual element antenna array (VEAA) and transmitting data collaboratively in a synchronous manner through a high-gain mainlobe (ML) beam. The goal is to optimize the deployment locations of UAVs in the VEAA and excitation current weights for performing CB transmissions considering the energy cost for UAV deployment. Accordingly, we formulate an Energy-Efficient Communication Multi-objective Optimization Problem (EECMOP) to jointly maximize the transmission rate and minimize the maximum sidelobe level (SLL) as well as UAV energy consumption. Then, we propose an Enhanced Multi Objective Ant Lion Optimizer (EMOALO) algorithm which incorporates a chaos theory to develop chaotic initialization and adjustable mode operators for solving the problem. Simulation results demonstrate the effectiveness of the EMOALO algorithm in improving energy efficiency for UAV-assisted communication networks.

Split learning (SL) is a collaborative learning framework, which can train an artificial intelligence (AI) model between a device and an edge server by splitting the AI model into a device-side model and a server-side model at a cut layer. The existing SL approach conducts the training process sequentially across devices, which incurs significant training latency especially when the number of devices is large. In this paper, we design a novel SL scheme to reduce the training latency, namedCluster-basedParallelSL(CPSL) which conducts model training in a “first-parallel-then-sequential” manner. Specifically, the CPSL is to partition devices into several clusters, parallelly train device-side models in each cluster and aggregate them, and then sequentially train the whole AI model across clusters, thereby parallelizing the training process and reducing training latency. Furthermore, we propose a resource management algorithm to minimize the training latency of CPSL considering device heterogeneity and network dynamics in wireless networks. This is achieved by stochastically optimizing the cut layer selection, device clustering, and radio spectrum allocation. The proposed two-timescale algorithm can jointly make the cut layer selection decision in a large timescale and device clustering and radio spectrum allocation decisions in a small timescale. Extensive simulation results on non-independent and identically distributed data demonstrate that the proposed solution can greatly reduce the training latency as compared with the existing SL benchmarks, while adapting to network dynamics.

In this paper, we investigate digital twin (DT)-assisted adaptive deep neural network (DNN) inference in the Industrial Internet of Things (IIoT). We consider a scenario that an edge server has a full-size DNN for high-accuracy inference, while an IIoT device has a lightweight DNN for fast on-device inference. The IIoT device generates computing tasks, such as object recognition, to be processed by DNN. For each task, a local controller at the network edge determines whether or not to offload the task to the edge server before it enters each layer of the lightweight DNN. The objective is to find the task offloading point that maximizes a utility including delay, inference accuracy, and on-device energy consumption. To achieve this objective, we propose an online DT-assisted task offloading scheme, which exploits DTs to capture the task processing status at the IIoT device and the workload at the edge server. Simulation results demonstrate the excellent performance of the proposed adaptive DT-assisted DNN inference on delay, inference accuracy, and on-device energy consumption.

In this paper, we present a digital twin (DT)-assisted adaptive video streaming scheme to enhance personalized quality-of-experience (PQoE). Since PQoE models are user-specific and time-varying, existing schemes based on universal and time-invariant PQoE models may suffer from performance degradation. To address this issue, we first propose a DT-assisted PQoE model construction method to obtain accurate user-specific PQoE models. Specifically, user DTs (UDTs) are respectively constructed for individual users, which can acquire and utilize users' data to accurately tune PQoE model parameters in real time. Next, given the obtained PQoE models, we formulate a resource management problem to maximize the overall long-term PQoE by taking the dynamics of users' locations, video requests, and buffer statuses into account. To solve this problem, a deep reinforcement learning algorithm is developed to jointly determine segment version selection, and communication and computing resource allocation. Simulation results on the real-world dataset demonstrate that the proposed scheme can effectively enhance PQoE compared with benchmark schemes.

This paper presents a novel approach for computing resource management of edge servers in vehicular networks based on digital twins and artificial intelligence (AI). Specifically, we construct two-tier digital twins tailored for vehicular networks to capture networking-related features of vehicles and edge servers. By exploiting such features, we propose a two-stage computing resource allocation scheme. First, the central controller periodically generates reference policies for real-time computing resource allocation according to the network dynamics and service demands captured by digital twins of edge servers. Second, computing resources of the edge servers are allocated in real time to individual vehicles via low-complexity matching-based allocation that complies with the reference policies. By leveraging digital twins, the proposed scheme can adapt to dynamic service demands and vehicle mobility in a scalable manner. Simulation results demonstrate that the proposed digital twin-driven scheme enables the vehicular network to support more computing tasks than benchmark schemes.

In this paper, we study Space-Air-Ground integrated Vehicular Network (SAGVN), and propose an online control framework to dynamically slice the SAG spectrum resource for isolated vehicular services provisioning. In particular, at a given time slot, the system makes online decisions on the request admission and scheduling, UAV dispatching, and resource slicing for different services. To characterize the impact of those parameters, we construct a time-averaged queue stability criteria by taking queue backlogs of all services into consideration, and formulate a system revenue function which incorporates the time-averaged system throughput and UAV dispatching cost. The objective is to maximize the system revenue while stabilizing the time-averaged queue, which falls into the scope of Lyapunov optimization theory. By bounding the drift-plus-penalty, the original problem can be decoupled into four independent subproblems, each of which is readily solved. The merits of our control framework are three-fold: 1) the system is able to admit and process as many requests as possible (i.e., maximizing the time-averaged throughput); 2) the time-averaged UAV dispatching cost is minimized; and 3) service queues are stabilized in the long-term. Extensive simulations are carried out, and the results demonstrate that the control framework can effectively achieve the system revenue maximization and queueing stabilization. Moreover, it can balance the trade-off among system throughput, UAV dispatching cost, and queueing states via parameter tuning. Compared with the fixed slicing, our dynamic slicing can react to the vehicular environment rapidly and achieve an average 26% of throughput improvement.

To support the explosive content demands from multifarious services and applications, cache-enabled satellite-terrestrial integrated networks (STINs) are envisioned as a key enabler to reduce the content delivery delay and alleviate the backhaul pressure. In this paper, we investigate the joint optimization of cache placement and content delivery in the STIN to minimize the long-term overall content delivery delay. Considering that cache placement and content delivery are interrelated and affected by network dynamics in terms of satellite movement and random content requests, the joint optimization problem is formulated as a sequential decision making problem by leveraging a Markov decision process. We propose a hierarchical deep Q learning (HDQL) algorithm by leveraging two independent deep neural networks to learn the cache placement and content delivery policies with small action space and low time complexity. Simulation results demonstrate that the proposed HDQL algorithm outperforms the benchmark algorithms in terms of content delivery delay in the STINs.

Space-ground integrated vehicular networks (SGIVNs) are envisioned as a promising architecture to support multifarious vehicular services with enhanced network flexibility and reliability. Access mode selection (AMS) is of capital importance in the SGIVN for the ingenious cooperation among different network segments to exploit their complementary advantages. In this paper, we investigate the AMS problem for vehicles in the SGIVN by taking distinct features of satellite networks (long propagation delay) and terrestrial networks (frequent handover) into account. In light of the high vehicle/satellite mobility and dynamic data packet arrivals, we formulate a stochastic integer programming problem of sequential AMS to maximize vehicles' long-term data rate. To cope with the time-varying network dynamics, we leverage a Markov decision process framework to model the evolution of vehicle states. For the special case with known stochastic model of data packet arrivals, we transform the problem into a linear programming problem that can be solved with low complexity. For the general case without the data packet arrival model, we propose a reinforcement learning-based algorithm to make adaptive AMS decisions to keep pace with network dynamics. Simulation results demonstrate that the proposed algorithm outperforms benchmark algorithms in terms of data rate under different data packet arrival patterns and vehicle velocities.

To support multifarious vehicular services with differentiated quality-of-service (QoS) requirements, space-air-ground integrated vehicular networks (SAGVNs) are envisioned as a promising solution to provide global network connectivity, enhance network flexibility, and improve network reliability. In this paper, we investigate cooperative content delivery in the SAGVN, where vehicular content requests can be simultaneously served by multiple access points (APs) in space, aerial, and terrestrial networks. In specific, a joint optimization problem of vehicle-to-AP association, bandwidth allocation, and content delivery ratio, referred to as the ABC problem, is formulated to minimize the overall content delivery delay while satisfying vehicular QoS requirements. To address the tightly-coupled optimization variables, we propose a load- and mobility-aware ABC (LMA-ABC) scheme to solve the joint optimization problem as follows. We first decompose the ABC problem to optimize the content delivery ratio. Then the impact of bandwidth allocation on the achievable delay performance is analyzed, and an effect of diminishing delay performance gain is revealed. Based on the analysis results, the LMA-ABC scheme is designed with the consideration of user fairness, load balancing, and vehicle mobility. Simulation results demonstrate that the proposed LMA-ABC scheme can significantly reduce the cooperative content delivery delay comparing to the benchmark schemes.

Drone cell (DC) is envisioned to enable the dynamic service provisioning for radio access networks (RANs), in response to the spatial and temporal unevenness of user traffic. In this article, we propose a hierarchical deep reinforcement learning (DRL)-based multi-DC trajectory planning and resource allocation (HDRLTPRA) scheme for high-mobility users. The objective is to maximize the accumulative network throughput while satisfying user fairness, DC power consumption, and DC-to-ground link quality constraints. To address the high uncertainties of the environment, we decouple the multi-DC TPRA problem into two hierarchical subproblems, i.e., the higher level global trajectory planning (GTP) subproblem and the lower level local TPRA (LTPRA) subproblem. First, the GTP subproblem is to address trajectory planning for multiple DCs in the RAN over a long time period. To solve the subproblem, we propose a multiagent DRL-based GTP (MARL-GTP) algorithm in which the nonstationary state space caused by the multi-DC environment is addressed by the multiagent fingerprint technique. Second, based on the GTP results, each DC solves the LTPRA subproblem independently to control the movement and transmit power allocation based on the real-time user traffic variations. A deep deterministic policy gradient (DEP)-based LTPRA (DEP-LTPRA) algorithm is then proposed to solve the LTPRA subproblem. With the two algorithms addressing both subproblems at different decision granularities, the multi-DC TPRA problem can be resolved by the HDRLTPRA scheme. Simulation results show that 40% network throughput improvement can be achieved by the proposed HDRLTPRA scheme over the nonlearning-based TPRA scheme.

In this paper, we investigate a radio access network (RAN) slicing problem for Internet of vehicles (IoV) services with different quality of service (QoS) requirements, in which multiple logically-isolated slices are constructed on a common roadside network infrastructure. A dynamic RAN slicing framework is presented to dynamically allocate radio spectrum and computing resource, and distribute computation workloads for the slices. To obtain an optimal RAN slicing policy for accommodating the spatial-temporal dynamics of vehicle traffic density, we first formulate a constrained RAN slicing problem with the objective to minimize long-term system cost. This problem cannot be directly solved by traditional reinforcement learning (RL) algorithms due to complicatedcoupled constraintsamong decisions. Therefore, we decouple the problem into a resource allocation subproblem and a workload distribution subproblem, and propose atwo-layer constrainedRL algorithm, namedResourceAllocation andWorkload diStribution (RAWS) to solve them. Specifically, anouter layerfirst makes the resource allocation decision via an RL algorithm, and then aninner layermakes the workload distribution decision via an optimization subroutine. Extensive trace-driven simulations show that the RAWS effectively reduces the system cost while satisfying QoS requirements with a high probability, as compared with benchmarks.

Collaboration among industrial Internet of Things (IoT) devices and edge networks is essential to support computation-intensive deep neural network (DNN) inference services, which require low delay and high accuracy. Sampling rate adaption, which dynamically configures the sampling rates of industrial IoT devices according to network conditions, is the key in minimizing the service delay. In this article, we investigate the collaborative DNN inference problem in industrial IoT networks. To capture the channel variation and task arrival randomness, we formulate the problem as a constrained Markov decision process (CMDP). Specifically, sampling rate adaption, inference task offloading, and edge computing resource allocation are jointly considered to minimize the average service delay while guaranteeing the long-term accuracy requirements of different inference services. Since CMDP cannot be directly solved by general reinforcement learning (RL) algorithms due to the intractable long-term constraints, we first transform the CMDP into an MDP by leveraging the Lyapunov optimization technique. Then, a deep RL-based algorithm is proposed to solve the MDP. To expedite the training process, an optimization subroutine is embedded in the proposed algorithm to directly obtain the optimal edge computing resource allocation. Extensive simulation results are provided to demonstrate that the proposed RL-based algorithm can significantly reduce the average service delay while preserving long-term inference accuracy with a high probability.

In this article, we investigate a computing task scheduling problem in space-air-ground integrated network (SAGIN) for delay-oriented Internet of Things (IoT) services. In the considered scenario, an unmanned aerial vehicle (UAV) collects computing tasks from IoT devices and then makes online offloading decisions, in which the tasks can be processed at the UAV or offloaded to the nearby base station or the remote satellite. Our objective is to design a task scheduling policy that minimizes offloading and computing delay of all tasks given the UAV energy capacity constraint. To this end, we first formulate the online scheduling problem as an energy-constrained Markov decision process (MDP). Then, considering the task arrival dynamics, we develop a novel deep risk-sensitive reinforcement learning algorithm. Specifically, the algorithm evaluates the risk, which measures the energy consumption that exceeds the constraint, for each state and searches the optimal parameter weighing the minimization of delay and risk while learning the optimal policy. Extensive simulation results demonstrate that the proposed algorithm can reduce the task processing delay by up to 30% compared to probabilistic configuration methods while satisfying the UAV energy capacity constraint.

Accurate cellular traffic load prediction is a pre-requisite for efficient and automatic network planning and management. Considering diverse users' activities at different locations and times, it is technically challenging to characterize the network resource demands at different time scales via traditional prediction methods. In this paper, we propose to combine the long short-term memory (LSTM) and Gaussian process regression (GPR) to achieve accurate single-cell level cellular traffic prediction, using the open Milan cellular traffic dataset provided by Telecom Italia. Firstly, the dominant periodic components of the cellular data are extracted, and then the small components are fed to the LSTM network. To further improve the prediction accuracy, GPR is used to recover the residual components. Extensive experiments are conducted based on the dataset, and it is shown that the proposed LSTM-GPR scheme outperforms the benchmark schemes, especially for a relatively long time and burst traffic prediction.

In this paper, we propose an online control frame-work to dynamically slice the network resource for isolated service provisioning in Space-Air-Ground integrated Vehicular Network (SAGVN). In particular, at a given time slot, the system makes online decisions on the request admission and scheduling, UAV dispatching, and resource slicing for different services. To characterize the impact of those parameters, we construct a time-averaged queue stability criteria by taking queue backlogs of all services into consideration, and formulate a system revenue function which incorporates the time-averaged system throughput and UAV dispatching cost. The objective is to maximize the system revenue while stabilizing the time-averaged queue, which can be achieved via the Lyapunov optimization theory. By bounding the drift-plus-penalty, the problem then can be decoupled into four independent subproblems, which are readily solved. The merits of our control framework are three-fold: 1) the system can admit and process as many requests as possible; 2) the time-averaged UAV dispatching cost is minimized; and 3) service queues can be stabilized over time. Extensive simulations are carried out, and the results demonstrate that the control framework can effectively achieve the system revenue maximization and queueing stabilization. Moreover, it can balance the trade-off among system throughput, UAV dispatching cost, and queueing states via parameter tuning.

In this article, we investigate the UAV-aided edge caching to assist terrestrial vehicular networks in delivering high-bandwidth content files. Aiming at maximizing the overall network throughput, we formulate a joint caching and trajectory optimization (JCTO) problem to make decisions on content placement, content delivery, and UAV trajectory simultaneously. As the decisions interact with each other and the UAV energy is limited, the formulated JCTO problem is intractable directly and timely. To this end, we propose a deep supervised learning scheme to enable intelligent edge for real-time decision-making in the highly dynamic vehicular networks. In specific, we first propose a clustering-based two-layered (CBTL) algorithm to solve the JCTO problem offline. With a given content placement strategy, we devise a time-based graph decomposition method to jointly optimize the content delivery and trajectory design, with which we then leverage the particle swarm optimization (PSO) algorithm to further optimize the content placement. We then design a deep supervised learning architecture of the convolutional neural network (CNN) to make fast decisions online. The network density and content request distribution with spatio-temporal dimensions are labeled as channeled images and input to the CNN-based model, and the results achieved by the CBTL algorithm are labeled as model outputs. With the CNN-based model, a function which maps the input network information to the output decision can be intelligently learnt to make timely inference and facilitate online decisions. We conduct extensive trace-driven experiments, and our results demonstrate both the efficiency of CBTL in solving the JCTO problem and the superior learning performance with the CNN-based model.

Due to the versatile networking capability, space- air-ground integrated network (SAGIN) becomes a prominent future architecture to support the ever- increasing Internet of Things (IoT) applications. In this paper, we investigate the IoT task offloading under an SAGIN scenario where multiple IoT devices generate computing tasks to be processed. We adopt an unmanned aerial vehicle (UAV) to fly along a given trajectory to collect the tasks of IoT devices within the coverage area, and then makes the online offloading decision, i.e., processing locally, or offloading to the nearby base station or the far-away satellite. However, due to the constrained energy resources committed by UAV and the uncertainty of the system dynamics, designing an efficient computation task offloading algorithm is challenging. This dynamic scheduling problem is formulated as a constrained Markov decision process (CMDP), considering the stochastic channel conditions, UAV coverage, energy consumption, and task queue backlogs. By exploiting the stationary stochastic feature of the CMDP, the problem can be solved by the linear programming to find a stochastic policy. Simulation results demonstrate that the proposed computation offloading scheme can significantly reduce IoT task processing delay as compared to other benchmarks.

Internet of Things (IoT) computing offloading is a challenging issue, especially in remote areas where common edge/cloud infrastructure is unavailable. In this paper, we present a space-air-ground integrated network (SAGIN) edge/cloud computing architecture for offloading the computation-intensive applications considering remote energy and computation constraints, where flying unmanned aerial vehicles (UAVs) provide near-user edge computing and satellites provide access to the cloud computing. First, for UAV edge servers, we propose a joint resource allocation and task scheduling approach to efficiently allocate the computing resources to virtual machines (VMs) and schedule the offloaded tasks. Second, we investigate the computing offloading problem in SAGIN and propose a learning-based approach to learn the optimal offloading policy from the dynamic SAGIN environments. Specifically, we formulate the offloading decision making as a Markov decision process where the system state considers the network dynamics. To cope with the system dynamics and complexity, we propose a deep reinforcement learning-based computing offloading approach to learn the optimal offloading policy on-the-fly, where we adopt the policy gradient method to handle the large action space and actor-critic method to accelerate the learning process. Simulation results show that the proposed edge VM allocation and task scheduling approach can achieve near-optimal performance with very low complexity and the proposed learning-based computing offloading algorithm not only converges fast but also achieves a lower total cost compared with other offloading approaches.