The rise of pre-trained foundation models (FMs) has popularized the trend of fine-tuning FMs to fit downstream tasks, while Federated Learning (FL) has become the de-facto approach for training distributed data with privacy-preservation. However, fine-tuning FMs in FL faces overwhelming overheads due to its bulky nature. While freezing parameters in FM have the potential to accelerate FL training, existing freezing strategies statically freeze parameters on specified or already converged layers, incur severe accuracy degradation, and resource-inefficiency in heterogeneous environments. In this paper, we propose AFedLF, an adaptive freezing framework for FM in FL, to accelerate its wall-clock time for convergence without losing its final accuracy. However, this poses great challenges, as different freezing strategies lead to different accuracy gains and time overheads, while unfreezing more layers may bring marginal accuracy gains but significant time overheads. To address this challenge, AFedLF mathematically establishes a correlation between the freezing strategy and the accuracy gain and time overhead, and allocates adaptive freezing strategies to clients, based on our insight that unfreezing more layers on devices with strong computation and communication capabilities helps improve resource efficiency. Besides, AFedLF incorporates our well-designed intermediate result caching scheme with constant approximation ratios utilizing the limited storage capacity on mobile devices to cache intermediate results to skip forward propagation, further saving wall-clock time. Finally, we implemented AFedLF using an open-source FL benchmark, and extensive trace-driven experimental results showed that AFedLF accelerates wall-clock time by up to 6.1× compared to state-of-the-art solutions, without sacrificing accuracy.

Ensuring the quality of service (QoS) in the Space-Air-Ground Integrated Network (SAGIN) requires implementing appropriate service function chain (SFC) embedding strategies to provide reliable general services and AI services. However, the complicated topology, limited infrastructure resources, and the heterogeneity in hardware and software in SAGIN hinder existing reliable embedding solutions from delivering high QoS. To address the challenges, this paper takes into account the heterogeneous reliable-aware SFC embedding (HRSE) problem in SAGIN. Specifically, we formulate the HRSE problem as a nonlinear integer programming problem that is NP-hard. Subsequently, an algorithm called HRSFCEA, which consists of deep reinforcement learning (DRL) and greedy rules, is proposed to tackle the challenges. Initially, leveraging the autonomous decision-making capability of DRL, SFC embedding is executed to select backups efficiently. The proposed algorithm reduces frequent trial and error by modifying its actions. Next, the paper gives the algorithm analysis including the complexity and unreliable environment. Finally, simulation results based on real-world datasets demonstrate that the DRL scheme converges to high rewards with minimal differences across varying numbers of hidden layer neurons. The proposed algorithm can improve the acceptance rate by up to 82% while incurring no more than 5% additional cost compared to the state-of-art algorithms.

Deep neural networks (DNNs) have demonstrated exceptional performance, leading to diverse applications across various mobile devices (MDs). Considering factors like portability and environmental sustainability, an increasing number of MDs are adopting energy harvesting (EH) techniques for power supply. However, the computational intensity of DNNs presents significant challenges for their deployment on these resource-constrained devices. Existing approaches often employ DNN partition or offloading to mitigate the time and energy consumption associated with running DNNs on MDs. Nonetheless, existing methods frequently fall short in accurately modeling the execution time of DNNs, and do not consider to use thread allocation for further latency and energy consumption optimization. To solve these problems, we propose a dynamic DNN partition and thread allocation method to optimize the latency and energy consumption of running DNNs on EH-enabled MDs. Specifically, we first investigate the relationship between DNN inference latency and allocated threads and establish an accurate DNN latency prediction model. Based on the prediction model, a DRL-based DNN partition (DDP) algorithm is designed to find the optimal partitions for DNNs. A thread allocation (TA) algorithm is proposed to reduce the inference latency. Experimental results from our test-bed platform demonstrate that compared to four benchmarking methods, our scheme can reduce DNN inference latency and energy consumption by up to 37.3% and 38.5%.

As key enabling technologies for 5G, edge computing pushes computing resources to the edge close to IoT users, enabling low-latency services. Meanwhile, network function virtualization abstracts complex services into service function chains (SFCs) consisting of multiple virtualized network functions (VNFs), significantly simplifying service management. However, the highly dynamic traffic in edge environments may cause the routing configurations of SFC requests and the resource configurations of VNFs to become outdated. The outdated configurations may result in VNF load imbalance or overload, severely disrupting service availability and degrading user quality of service (QoS). Therefore, this paper studies the novel joint optimization of SFC routing update and elastic resource configuration problem, aiming to minimize the resource leasing cost of the service provider, while subject to multi-resource capacity and update delay constraints. Considering all these characteristics, we formalize this problem as an integer linear programming and prove its NP-hardness. To tackle this problem, we propose a rounding-based routing update and resource configuration algorithm to find cost-effective routing paths and configuration types for requests and VNFs. Further, we design a backtracking-based greedy improvement algorithm that upgrades VNF resource configurations and adjusts request routing paths to accommodate more requests, thus boosting network throughput. Extensive simulation results show that compared with state-of-the-art solutions, our scheme can reduce the leasing cost by 17.8%, while improving network throughput and meeting user QoS.

As a key enabler for 6G and symbiotic IoT applications, edge computing brings computing resources closer to end users, ensuring low-latency services, while network function virtualization (NFV) enables flexible service deployment by abstracting complex functionalities into service function chains (SFCs) composed of virtual network functions (VNFs). However, in dynamic edge environments, traffic changes and frequent network failures can lead to outdated routing configurations, resulting in load imbalance, network congestion, and Quality of Service (QoS) degradation. These issues will damage the robustness and real-time response capability required for symbiotic IoT systems. Although existing work separately optimizes VNF assignment or SFC routing for real-time and robust SFC updates, it fails to jointly optimize both, leading to potential VNF overload and throughput degradation. Therefore, this article studies the joint optimization of the VNF assignment and SFC routing problem, considering limited resource capacity, robustness and delay requirements, and bounded migration costs, aiming to maximize the network throughput. To capture all these characteristics, we formalize the problem as integer linear programming (ILP) and prove its NP-hardness. To tackle this problem, we propose a QoS-aware VNF assignment and routing update algorithm that first relaxes the ILP to linear programming (LP), and then randomly rounds the fractional solution obtained by solving LP to obtain a nearly optimal integer solution. Furthermore, we propose a backtracking-based greedy improvement algorithm, which greedily adjusts VNF assignments and request paths to accommodate more requests and satisfy all constraints. Extensive simulation results show that compared to state-of-the-art solutions, the proposed method can improve throughput by 22.98% while meeting user QoS.

As smart mobile applications increasingly demand timely situational awareness and energy efficiency, the Age of Information (AoI) metric plays a vital role in maintaining data freshness. This need is further supported by the End-Edge-Cloud Computing (EECC) paradigm, which enhances application performance by facilitating task offloading to the edge or the cloud. However, existing AoI optimization solutions focus solely on task offloading, often neglecting critical aspects such as system resource allocation and energy efficiency, which can lead to resource waste, increased energy consumption, compromised Quality of Service (QoS), and system performance degradation. Therefore, this paper investigates the joint optimization of task offloading, communication and computing resource allocation in EECC systems, aiming to minimize AoI and energy consumption under constraints of deadlines and capacity constraints. To address this problem, we divide the decision space into multiple non-intersecting decision areas based on the characteristics of the studied problem and design a task offloading and resource allocation algorithm based on slow-movement particle swarm optimization (SPSO) to handle each decision area individually. In the algorithm design, we customize the position, velocity, update rules, and fitness function for the optimization problem. Finally, extensive simulation-based and testbed experiment results show that the proposed algorithm can save up to 14.56% of energy consumption, shorten AoI by up to 27.80%, and improve utility (weighted sum of AoI and energy consumption) by up to 15.89% compared with existing algorithms.

Federated learning (FL) enables privacy-preserving distributed machine learning by training models on edge client devices using their local data without revealing their raw data. In edge environments, various applications require different neural network models, making it crucial to perform joint training of multiple models on edge devices, known as multi-task FL. While existing multi-task FL approaches enhance resource utilization on edge devices through adaptive resource configuration or client selection, optimizing either of these aspects alone may lead to suboptimality. Therefore, in this paper, we explore a joint client selection and resource configuration method called JCSRC for multi-task FL, aiming to maximize energy efficiency in environments with limited computation and communication resources and heterogeneous client devices. Firstly, we formalize this problem as a mixed-integer nonlinear programming problem considering all these characteristics and prove its NP-hardness. To address this problem, we first design a multi-agent reinforcement learning (MARL)-based client selection method that selects appropriate clients for each task to train their models. The MARL method makes client selection decisions based on the clients’ data quality, energy efficiency, communication, and computation capacity to ensure fast convergence and energy efficiency. Then, we design a particle swarm optimization (PSO)-based resource configuration scheme that configures appropriate computation and bandwidth resources for each task on each client. The PSO scheme makes resource configuration decisions based on theoretically derived optimal CPU frequency and bandwidth to achieve high energy efficiency. Finally, we carry out extensive simulations and testbed-based experiments to validate our proposed JCSRC. The results demonstrate that, in comparison to state-of-the-art solutions, JCSRC can save energy consumption by up to 59% to achieve the target accuracy.

The prevalence of the pre-training & fine-tuning paradigm enables machine learning models to quickly adapt to various downstream tasks by fine-tuning pre-trained foundation models (FMs), greatly facilitating various IoT applications that rely on model inference in dynamic edge serverless environments. Efficiently dispatching inference requests and configuring instances to batch inference requests can significantly enhance resource efficiency. However, existing serverless inference solutions are tailored for traditional models, make coarse-grained request dispatching and instance configuration decisions, fail to exploit the shared model backbone characteristics of the FM and capture delayed rewards in dynamic environments, and ignore communication latency between edge sites, resulting in high costs and constraint violations. In this paper, we leverage our insight that fine-grained batch inference requests can effectively exploit the shared model backbone feature of FM to save monetary costs. We propose an algorithm that incorporates deep reinforcement learning (DRL) and expert intervention for fine-grained request dispatching and instance configuration, where the DRL component outputs fractional solutions as guidance, while the expert intervention module integrates our insights—batching reduces monetary costs at the expense of increased inference latency, whereas higher configurations shorten inference latency. This module rounds fractional solutions and adjusts instance configurations to search for optimal solutions while satisfying constraints, with theoretical guarantees rigorously proved. Finally, we conducted our experiments on an OpenFaas-based platform and simulator, and extensive trace-driven evaluation results show that ExpertDRL can save costs by up to 85.14% and improve request acceptance ratio by up to 26.93%, compared to the state-of-the-art solution.

Mobile-edge computing (MEC) has received wide attention recently due to its efficacy in alleviating the computation stress of mobile devices (MDs), which is realized by offloading workloads from MD users to nearby edge servers (ESs). Prior work has studied related task offloading and scheduling problems and proposed many approaches. However, none of these approaches considers the reliability issue in MEC systems which may suffer soft errors during task execution as well as bit errors during task offloading simultaneously. Targeting optimization on a multi-user MEC system, in this article we investigate the task offloading and scheduling problem of maximizing system quality of experience (QoE) under a certain reliability requirement. With the consideration of the combinatorial nature of this problem, we propose to decompose the original problem into i) a task-to-ES assignment problem with fixed task offloading decision, for satisfying system reliability constraint, ii) a computing resource allocation problem with fixed task offloading and assignment decisions, for maximizing system QoE, and iii) a task offloading optimization problem to find the best offloading decision that achieves the maximum QoE under the reliability constraint using our task assignment and resource allocation methods. In order to solve these sub-problems, we further design a reliability-aware task-to-ES assignment algorithm, a QoE-optimum resource allocation algorithm, and a binary particle swarm optimization based task offloading algorithm. We perform extensive simulations and testbed experiments to validate the efficacy of the proposed scheme. Simulation and testbed results show that the proposed scheme greatly outperforms four benchmark approaches and it achieves up to 63.2% and 43.1% increase in the average QoE (quantified by offloading utility), respectively.

Serverless edge computing uses lightweight containers to execute IoT applications on-demand, enhancing resource utilization efficiency at the network edge. However, cold-start latency, occurring when a container is instantiated from scratch, degrades IoT service responsiveness. Effective mitigation strategies are needed to maintain service quality. Current container caching strategies often fail to account for diverse workloads and predictable request patterns. This paper introduces a dynamic container pre-warm model that adjusts container states based on predicted future requests, mitigating cold-start overhead. Our approach employs a real-time algorithm for container instantiation and termination, balancing latency reduction and resource conservation. Inte-grating predictive analytics with adaptive resource management, we evaluate our model using a trace-driven simulation framework with real-world data. Our experiments show that the proposed strategy reduces cold starts by up to 60 % and improves average response times by 45 % compared to state-of-the-art methods. Additionally, our approach adapts to varying workloads, providing a scalable and robust solution for dynamic edge environments.

As a key technology of 5G, network function virtualization enables each monolithic service to be divided into microservices, facilitating their deployment and management in edge environments. One of the most critical issues in 5G is how to support dynamically arriving mission-critical services with low-latency and high-reliability requirements in distributed edge environments. However, most existing works focus on how to provide reliable services without considering latency, and their heuristics struggle to cope with high-dimensional constraints and complex environments with heterogeneous infrastructure and services. In this paper, we propose a SafeDRL algorithm to resource-efficiently support these dynamically arriving services while meeting their reliability and latency requirements. Specifically, we first formulate the problem as an integer nonlinear programming and prove its NP-hardness. To tackle this problem, our SafeDRL algorithm captures delayed rewards in dynamic environments by reinforcement learning, and corrects constraint violations with high-quality feasible solutions based on expert intervention, and prunes unnecessary backup instances for optimality. The algorithm is proved to have a bounded approximation ratio in general cases. Extensive trace-driven simulations show that, compared with the state-of-the-art solution, SafeDRL can save resource costs by up to 49.32% and improve the service acceptance ratio by up to 55% with acceptable execution time.

Propelled by recent advances in Mobile Edge Computing (MEC) and the Internet of Things (IoT), the digital twin technique has been envisioned as a de-facto driving force to bridge the virtual and physical worlds through creating digital portrayals of physical objects. In virtue of the flourishing of edge intelligence and abundant IoT data, data-driven modelling facilitates the implementation and maintenance of digital twins, where simulations of physical objects are usually performed based on Deep Neural Networks (DNNs). A significant advantage of adopting digital twins is to enable decisive prediction on the behaviours of objects in near future without waiting for that really happen. To provide accurate predictions, it is vital to keep each digital twin synchronized with its physical object in real-time. However, it is challenging to maintain the real-time synchronization between a digital twin and its physical object due to the dynamics of physical objects and sensing data drift over time, i.e., the live data from a physical object diverge from the model training data of its digital twin. To address this critical issue, continual learning is a promising solution to retrain models of digital twins incrementally. In this paper, we investigate digital twin synchronization issues via continual learning in an MEC environment, with the aim to maximize the total utility gain, i.e., the enhanced model accuracy. We study two novel optimization problems: the static digital twin synchronization problem per time slot and the dynamic digital twin synchronization problem for a finite time horizon. We first formulate an Integer Linear Program (ILP) solution for the static digital twin synchronization problem when the problem size is small; otherwise, we develop a randomized approximation algorithm at the expense of bounded resource violations for it. We also devise a deterministic approximation algorithm with guaranteed performance for a special case of the static digital twin synchronization problem. We thirdly consider the dynamic digital twin synchronization problem by proposing an efficient online algorithm for it. Finally, we evaluate the performance of the proposed algorithms for continuous digital twin synchronization through simulations. Simulation results show that the proposed algorithms are promising, outperforming counterpart benchmarks by no less than 13.2%, in terms of the total utility gain.

Traditional traffic classification methods struggle to identify emerging network traffic due to the need for model retraining, which hampers the real-time response of deployed edge devices. Furthermore, emerging network traffic samples are often scarce, traditional methods often treat a session as a single image, thereby overlooking essential structural features. These factors can result in poor generalization ability of the trained model. To overcome these challenges, we propose ConViTML (Convolutional Vision Transformer-based Meta-Learning), a real-time end-to-end network traffic classification framework that employs meta-learning to avoid model retraining. We propose a novel feature extraction network, Convolutional Visual Transformer (ConViT), merging Convolutional Neural Network (CNN) and Visual Transformer (ViT). ConViT can directly extract low-dimensional discriminative features containing basic and structural features of the session, which is vital for improving detection accuracy and accelerating convergence in a data-scarce environment. Furthermore, we employ a Packet-based Relation Network (PRN) to analyze the matching degree of support samples and query samples. Therefore, accurate classification in novel traffic identification tasks can be achieved with just a few labeled samples, eliminating extensive data collection and labeling operations. Finally, we replace various feature extractors and compare our approach with the classic meta-learning framework Relation Network (RelationNet). Extensive experimental results demonstrate that ConViTML outperforms others with various performance indicators.

Recent studies usually approach multi-label zeroshot learning (MLZSL) with visual-semantic mapping on spatial-class correlation, which can be computationally costly, and worse still, fails to capture fine-grained classspecific semantics. We observe that different channels may usually have different sensitivities on classes, which can correspond to specific semantics. Such an intrinsic channelclass correlation suggests a potential alternative for the more accurate and class-harmonious feature representations. In this paper, our interest is to fully explore the power of channel-class correlation as the unique base for MLZSL. Specifically, we propose a light yet efficient Multi-Label Multi-Layer Perceptron-based Encoder, dubbed (ML)2P-Encoder, to extract and preserve channel-wise semantics. We reorganize the generated feature maps into several groups, of which each of them can be trained independently with (ML)2P-Encoder. On top of that, a global groupwise attention module is further designed to build the multilabel specific class relationships among different classes, which eventually fulfills a novel Channel-Class Correlation MLZSL framework (C3-MLZSL)11Released code:github.com/simonzmliu/cvpr23_mlzsl. Extensive experiments on large-scale MLZSL benchmarks including NUS-WIDE and Open-Images-V4 demonstrate the superiority of our model against other representative state-of-the-art models.

The software-defined network (SDN) enabled mobile edge network greatly facilitates network resource management and promotes many emerging applications. However, user mobility may cause the SDN controller to set flow rules frequently, introduce additional flow setup latency, cause delay jitter, and undermine latency-sensitive services. Proactive flow setup is an effective way to eliminate flow setup latency, but existing work fails to maximize the flow setup hit ratio, a metric for evaluating the quality of proactive flow setup decisions, which is critical for latency-sensitive services. In this paper, we study how to proactively set flow rules to maximize the flow setup hit ratio under limited available network resources to eliminate the flow setup latency as much as possible. Then, we formalize the proactive flow setup problem as two integer linear programming problems under two typical routing strategies, default routing and dynamic routing. Both problems are proved to be NP-hard. To tackle these two problems, we propose a linear programming-based polynomial-time approximation algorithm for the default routing case and a greedy-based heuristic algorithm for the dynamic routing case. Extensive trace-driven experimental and simulation results verify that our algorithms can improve the flow setup hit ratio by up to 30.99% compared to existing solutions.

As a key enabling technology for 5G, network function virtualization abstracts services into software-based service function chains (SFCs), facilitating mission-critical services with high-reliability requirements. However, it is challenging to cost-effectively provide reliable SFCs in dynamic environments due to delayed rewards caused by future SFC requests, limited infrastructure resources, and heterogeneity in hardware and software reliability. Although deep reinforcement learning (DRL) can effectively capture delayed rewards in dynamic environments, its trial-and-error exploration in a vast solution space with massive infeasible solutions may lead to frequent constraint violations and traps in poor local optima. To address these challenges, we propose a RuleDRL algorithm that combines the capability of DRL to capture delayed rewards and the strength of rule-based schemes to explore high-quality solutions without violating constraints. Specifically, we first formulate the reliable SFC provision problem as an integer nonlinear programming problem, which is proven to be NP-hard. Then, we jointly design DRL and rule-based schemes that are coupled to make the final decision and establish a bounded approximation ratio in general cases. Extensive trace-driven simulations show that RuleDRL can save the total cost by up to 65.67% and improve the SFC acceptance ratio by up to 82%, compared to the state-of-the-art solution.

With the prosperity of cloud computing and video services, the demand for network resources has increased dramatically, leading to the remarkable growth in the amount of network energy consumption, a key factor restricting the development of data centers. Numerous existing works reduce network energy consumption by optimizing data transmission, but they ignore the energy consumption for data transmission preparation, such as activating devices and installing rules. In this paper, we jointly optimize device activation, rule installation and data transmission to minimize network energy consumption. Specifically, we first formulate the minimization problem of the energy consumption of device activation, rule installation, and data transmission. We then prove that it is NP-complete to get the optimal solution of the minimization problem, furthermore, we propose a heuristic algorithm to plan the path with minimum network energy consumption for each flow. The simulation results show that the energy consumption of our algorithm is close to the optimal solution solved by Gurobi, and our algorithm has lower complexity. Compared with the state-of-the-art algorithm, our algorithm always consumes less energy and has shorter flow completion time.

Summary Data transmission in current networks is usually associated with strict priority enforcement for the purpose of quality of service (QoS). Under the case that priority flow requests are injected into the network sequentially without the information of future flow request arrivals, it is a challenging to achieve network throughput maximization for on‐line flow requests under the joint constraints of the flow's priority, bandwidth demand, and resource capacity. Software Defined Networking (SDN) can effectively solve the flow scheduling equilibrium problem between the priority of dynamic flow requests and the maximization of network throughput. Therefore, in this paper, we study on‐line flow request admission in SDN, the goal of which is to maximize the network throughput under the constraints of critical network bandwidth resources, flow priority, and bandwidth demands. First, we present the concept of flow routing cost and profit and a model to characterize the cost of using link resources and routing paths. Then, we propose an efficient on‐line priority flow scheduling algorithm (OPFSA) to solve priority flow request scheduling problem and analyze the competitive ratio of OPFSA. Our on‐line algorithm can reach throughput within of the highest possible throughput that can be achieved by an off‐line algorithm, where n is the number of node in the network. Finally, experimental results demonstrate that compared with SHORTEST‐SC, our proposed algorithm can enhance the cumulative bandwidth about 9% and 40% when general network size is 30 and 170 nodes, respectively, and improve the throughput about 25% in Fat‐tree network when pod size is 4.