Using satellites to complete spectrum monitoring tasks can effectively receive and process electromagnetic spectrum signals emitted by radiation sources.However,due to the shortage of satellite storage,computing and n...Using satellites to complete spectrum monitoring tasks can effectively receive and process electromagnetic spectrum signals emitted by radiation sources.However,due to the shortage of satellite storage,computing and network resources,the intersatellite coordination is weak,and with the massive growth of spectrum data,the traditional cloud computing mode cannot meet the requirements of electromagnetic spectrum monitoring in terms of real-time,bandwidth,and security.We apply edge computing technology and deep learning technology to the satellite.Aiming at the problems of distributed satellite management and control,we propose a space-based distributed electromagnetic spectrum monitoring intelligent connected cloud-edge collaborative architecture SpaceEdge.SpaceEdge applies edge computing and artificial intelligence technology to space-based spectrum monitoring.SpaceEdge deploys intelligent monitoring algorithms to edge nodes to form edge intelligent satellite,and uses the cloud to uniformly manage and control heterogeneous edge satellite and monitor satellite resources.In addition,SpaceEdge can also adjust edge intelligent spectrum monitoring applications as needed to achieve effective coordination of inter-satellite algorithms and data to achieve the purpose of collaborative monitoring.Finally,SpaceEdge was experimentally verified,and the results proved the feasibility of SpaceEdge and can improve the timeliness and autonomy of the distributed satellite’s coordinated signal monitoring.展开更多
With the continuous development of network func-tions virtualization(NFV)and software-defined networking(SDN)technologies and the explosive growth of network traffic,the requirement for computing resources in the netw...With the continuous development of network func-tions virtualization(NFV)and software-defined networking(SDN)technologies and the explosive growth of network traffic,the requirement for computing resources in the network has risen sharply.Due to the high cost of edge computing resources,coordinating the cloud and edge computing resources to improve the utilization efficiency of edge computing resources is still a considerable challenge.In this paper,we focus on optimiz-ing the placement of network services in cloud-edge environ-ments to maximize the efficiency.It is first proved that,in cloud-edge environments,placing one service function chain(SFC)integrally in the cloud or at the edge can improve the utilization efficiency of edge resources.Then a virtual network function(VNF)performance-resource(P-R)function is proposed to repre-sent the relationship between the VNF instance computing per-formance and the allocated computing resource.To select the SFCs that are most suitable to deploy at the edge,a VNF place-ment and resource allocation model is built to configure each VNF with its particular P-R function.Moreover,a heuristic recur-sive algorithm is designed called the recursive algorithm for max edge throughput(RMET)to solve the model.Through simula-tions on two scenarios,it is verified that RMET can improve the utilization efficiency of edge computing resources.展开更多
In this paper,we propose a novel fuzzy matching data sharing scheme named FADS for cloudedge communications.FADS allows users to specify their access policies,and enables receivers to obtain the data transmitted by th...In this paper,we propose a novel fuzzy matching data sharing scheme named FADS for cloudedge communications.FADS allows users to specify their access policies,and enables receivers to obtain the data transmitted by the senders if and only if the two sides meet their defined certain policies simultaneously.Specifically,we first formalize the definition and security models of fuzzy matching data sharing in cloud-edge environments.Then,we construct a concrete instantiation by pairing-based cryptosystem and the privacy-preserving set intersection on attribute sets from both sides to construct a concurrent matching over the policies.If the matching succeeds,the data can be decrypted.Otherwise,nothing will be revealed.In addition,FADS allows users to dynamically specify the policy for each time,which is an urgent demand in practice.A thorough security analysis demonstrates that FADS is of provable security under indistinguishable chosen ciphertext attack(IND-CCA)in random oracle model against probabilistic polynomial-time(PPT)adversary,and the desirable security properties of privacy and authenticity are achieved.Extensive experiments provide evidence that FADS is with acceptable efficiency.展开更多
Nowadays,the rapid development of edge computing has driven an increasing number of deep learning applications deployed at the edge of the network,such as pedestrian and vehicle detection,to provide efficient intellig...Nowadays,the rapid development of edge computing has driven an increasing number of deep learning applications deployed at the edge of the network,such as pedestrian and vehicle detection,to provide efficient intelligent services to mobile users.However,as the accuracy requirements continue to increase,the components of deep learning models for pedestrian and vehicle detection,such as YOLOv4,become more sophisticated and the computing resources required for model training are increasing dramatically,which in turn leads to significant challenges in achieving effective deployment on resource-constrained edge devices while ensuring the high accuracy performance.For addressing this challenge,a cloud-edge collaboration-based pedestrian and vehicle detection framework is proposed in this paper,which enables sufficient training of models by utilizing the abundant computing resources in the cloud,and then deploying the well-trained models on edge devices,thus reducing the computing resource requirements for model training on edge devices.Furthermore,to reduce the size of the model deployed on edge devices,an automatic pruning method combines the convolution layer and BN layer is proposed to compress the pedestrian and vehicle detection model size.Experimental results show that the framework proposed in this paper is able to deploy the pruned model on a real edge device,Jetson TX2,with 6.72 times higher FPS.Meanwhile,the channel pruning reduces the volume and the number of parameters to 96.77%for the model,and the computing amount is reduced to 81.37%.展开更多
With the construction of the power Internet of Things(IoT),communication between smart devices in urban distribution networks has been gradually moving towards high speed,high compatibility,and low latency,which provi...With the construction of the power Internet of Things(IoT),communication between smart devices in urban distribution networks has been gradually moving towards high speed,high compatibility,and low latency,which provides reliable support for reconfiguration optimization in urban distribution networks.Thus,this study proposed a deep reinforcement learning based multi-level dynamic reconfiguration method for urban distribution networks in a cloud-edge collaboration architecture to obtain a real-time optimal multi-level dynamic reconfiguration solution.First,the multi-level dynamic reconfiguration method was discussed,which included feeder-,transformer-,and substation-levels.Subsequently,the multi-agent system was combined with the cloud-edge collaboration architecture to build a deep reinforcement learning model for multi-level dynamic reconfiguration in an urban distribution network.The cloud-edge collaboration architecture can effectively support the multi-agent system to conduct“centralized training and decentralized execution”operation modes and improve the learning efficiency of the model.Thereafter,for a multi-agent system,this study adopted a combination of offline and online learning to endow the model with the ability to realize automatic optimization and updation of the strategy.In the offline learning phase,a Q-learning-based multi-agent conservative Q-learning(MACQL)algorithm was proposed to stabilize the learning results and reduce the risk of the next online learning phase.In the online learning phase,a multi-agent deep deterministic policy gradient(MADDPG)algorithm based on policy gradients was proposed to explore the action space and update the experience pool.Finally,the effectiveness of the proposed method was verified through a simulation analysis of a real-world 445-node system.展开更多
Software-defined networking(SDN)enables the separation of control and data planes,allowing for centralized control and management of the network.Without adequate access control methods,the risk of unau-thorized access...Software-defined networking(SDN)enables the separation of control and data planes,allowing for centralized control and management of the network.Without adequate access control methods,the risk of unau-thorized access to the network and its resources increases significantly.This can result in various security breaches.In addition,if authorized devices are attacked or controlled by hackers,they may turn into malicious devices,which can cause severe damage to the network if their abnormal behaviour goes undetected and their access privileges are not promptly restricted.To solve those problems,an anomaly detection and access control mechanism based on SDN and neural networks is proposed for cloud-edge collaboration networks.The system employs the Attribute Based Access Control(ABAC)model and smart contract for fine-grained control of device access to the network.Furthermore,a cloud-edge collaborative Key Performance Indicator(KPI)anomaly detection method based on the Gated Recurrent Unit and Generative Adversarial Nets(GRU-GAN)is designed to discover the anomaly devices.An access restriction mechanism based on reputation value and anomaly detection is given to prevent anomalous devices.Experiments show that the proposed mechanism performs better anomaly detection on several datasets.The reputation-based access restriction effectively reduces the number of malicious device attacks.展开更多
Cloud storage and edge computing are utilized to address the storage and computational challenges arising from the exponential data growth in IoT.However,data privacy is potentially risky when data is outsourced to cl...Cloud storage and edge computing are utilized to address the storage and computational challenges arising from the exponential data growth in IoT.However,data privacy is potentially risky when data is outsourced to cloud servers or edge services.While data encryption ensures data confidentiality,it can impede data sharing and retrieval.Attribute-based searchable encryption(ABSE)is proposed as an effective technique for enhancing data security and privacy.Nevertheless,ABSE has its limitations,such as single attribute authorization failure,privacy leakage during the search process,and high decryption overhead.This paper presents a novel approach called the blockchain-assisted efficientmulti-authority attribute-based searchable encryption scheme(BEM-ABSE)for cloudedge collaboration scenarios to address these issues.BEM-ABSE leverages a consortium blockchain to replace the central authentication center for global public parameter management.It incorporates smart contracts to facilitate reliable and fair ciphertext keyword search and decryption result verification.To minimize the computing burden on resource-constrained devices,BEM-ABSE adopts an online/offline hybrid mechanism during the encryption process and a verifiable edge-assisted decryption mechanism.This ensures both low computation cost and reliable ciphertext.Security analysis conducted under the random oracle model demonstrates that BEM-ABSE is resistant to indistinguishable chosen keyword attacks(IND-CKA)and indistinguishable chosen plaintext attacks(INDCPA).Theoretical analysis and simulation results confirm that BEM-ABSE significantly improves computational efficiency compared to existing solutions.展开更多
As the penetration rate of renewable energy sources(RES)gradually increases,demand-side resources(DSR)should be fully utilized to provide flexibility and rapidly respond to real-time power supply-demand imbalance.Howe...As the penetration rate of renewable energy sources(RES)gradually increases,demand-side resources(DSR)should be fully utilized to provide flexibility and rapidly respond to real-time power supply-demand imbalance.However,scheduling a large number of DSR clusters will inevitably bring unbearable transmission delay,and computation delay,which in turn lead to lower response speeds.This paper examines flexibility scheduling of DSR clusters within a smart distribution network(SDN)in view of both kinds of delay.Building upon a SDN model,maximum schedulable flexibility of DSR clusters is first quantified.Then,a flexibility response curve is analyzed to reflect the effect of delay on flexibility scheduling.Aiming at reducing flexibility shortage brought by delay,we propose a modified flexibility scheduling strategy based on cloud-edge collaboration.Compared with traditional strategy,centralized optimization is replaced by distributed optimization to consider both economic efficiency and effect of delay.Besides,an offloading strategy is also formulated to decide optimal edge nodes and corresponding wired paths for edge computations.In a case study,we evaluate scheduled flexibility,operational cost,average delay and the chosen edge nodes for edge computations with traditional strategy and our proposed strategy.Evaluation results show the proposed strategy can significantly reduce the effect of delay on flexibility scheduling,and guarantee the optimality of operational cost to some extent.展开更多
Service recommendation provides an effective solution to extract valuable information from the huge and ever-increasing volume of big data generated by the large cardinality of user devices.However,the distributed and...Service recommendation provides an effective solution to extract valuable information from the huge and ever-increasing volume of big data generated by the large cardinality of user devices.However,the distributed and rich multi-source big data resources raise challenges to the centralized cloud-based data storage and value mining approaches in terms of economic cost and effective service recommendation methods.In view of these challenges,we propose a deep neural collaborative filtering based service recommendation method with multi-source data(i.e.,NCF-MS)in this paper,which adopts the cloud-edge collaboration computing paradigm to build recommendation model.More specifically,the Stacked Denoising Auto Encoder(SDAE)module is adopted to extract user/service features from auxiliary user profiles and service attributes.The Multiple Layer Perceptron(MLP)module is adopted to integrate the auxiliary user/service features to train the recommendation model.Finally,we evaluate the effectiveness of the NCF-MS method on three public datasets.The experimental results show that our proposed method achieves better performance than existing methods.展开更多
How to collaboratively offload tasks between user devices,edge networks(ENs),and cloud data centers is an interesting and challenging research topic.In this paper,we investigate the offoading decision,analytical model...How to collaboratively offload tasks between user devices,edge networks(ENs),and cloud data centers is an interesting and challenging research topic.In this paper,we investigate the offoading decision,analytical modeling,and system parameter optimization problem in a collaborative cloud-edge device environment,aiming to trade off different performance measures.According to the differentiated delay requirements of tasks,we classify the tasks into delay-sensitive and delay-tolerant tasks.To meet the delay requirements of delay-sensitive tasks and process as many delay-tolerant tasks as possible,we propose a cloud-edge device collaborative task offoading scheme,in which delay-sensitive and delay-tolerant tasks follow the access threshold policy and the loss policy,respectively.We establish a four-dimensional continuous-time Markov chain as the system model.By using the Gauss-Seidel method,we derive the stationary probability distribution of the system model.Accordingly,we present the blocking rate of delay-sensitive tasks and the average delay of these two types of tasks.Numerical experiments are conducted and analyzed to evaluate the system performance,and numerical simulations are presented to evaluate and validate the effectiveness of the proposed task offloading scheme.Finally,we optimize the access threshold in the EN buffer to obtain the minimum system cost with different proportions of delay-sensitive tasks.展开更多
Market participants can only bid with lagged information disclosure under the existing market mechanism,which can lead to information asymmetry and irrational market behavior,thus influencing market efficiency.To prom...Market participants can only bid with lagged information disclosure under the existing market mechanism,which can lead to information asymmetry and irrational market behavior,thus influencing market efficiency.To promote rational bidding behavior of market participants and improve market efficiency,a novel electricity market mechanism based on cloudedge collaboration is proposed in this paper.Critical market information,called residual demand curve,is published to market participants in real-time on the cloud side,while participants on the edge side are allowed to adjust their bids according to the information disclosure prior to closure gate.The proposed mechanism can encourage rational bids in an incentive-compatible way through the process of dynamic equilibrium while protecting participants’privacy.This paper further formulates the mathematical model of market equilibrium to simulate the process of each market participant’s strategic bidding behavior towards equilibrium.A case study based on the IEEE 30-bus system shows the proposed market mechanism can effectively guide bidding behavior of market participants,while condensing exchanged information and protecting privacy of participants.展开更多
With the increasing penetration of renewable energy generation,uncertainty and randomness pose great challenges for optimal dispatching in distribution networks.We propose a cloud-edge cooperative dispatching(CECD)met...With the increasing penetration of renewable energy generation,uncertainty and randomness pose great challenges for optimal dispatching in distribution networks.We propose a cloud-edge cooperative dispatching(CECD)method to exploit the new opportunities offered by Internet of Things(IoT)technology.To alleviate the huge pressure on the modeling and computing of large-scale distribution system,the method deploys edge nodes in small-scale transformer areas in which robust optimization subproblem models are introduced to address the photovoltaic(PV)uncertainty.Considering the limited communication and computing capabilities of the edge nodes,the cloud center in the distribution automation system(DAS)establishes a utility grid master problem model that enforces the consistency between the solution at each edge node with the utility grid based on the alternating direction method of multipliers(ADMM).Furthermore,the voltage constraint derived from the linear power flow equations is adopted for enhancing the operation security of the distribution network.We perform a cloud-edge system simulation of the proposed CECD method and demonstrate a dispatching application.The case study is carried out on a modified 33-node system to verify the remarkable performance of the proposed model and method.展开更多
With the extensive penetration of distributed renewable energy and self-interested prosumers,the emerging power market tends to enable user autonomy by bottom-up control and distributed coordination.This paper is devo...With the extensive penetration of distributed renewable energy and self-interested prosumers,the emerging power market tends to enable user autonomy by bottom-up control and distributed coordination.This paper is devoted to solving the specific problems of distributed energy management and autonomous bidding and peer-to-peer(P2P)energy sharing among prosumers.A novel cloud-edge-based We-Market is presented,where the prosumers,as edge nodes with independent control,balance the electricity cost and thermal comfort by formulating a dynamic household energy management system(HEMS).Meanwhile,the autonomous bidding is initiated by prosumers via the modified Stone-Geary utility function.In the cloud center,a distributed convergence bidding(CB)algorithm based on consistency criterion is developed,which promotes faster and fairer bidding through the interactive iteration with the edge nodes.Besides,the proposed scheme is built on top of the commercial cloud platform with sufficiently secure and scalable computing capacity.Numerical results show the effectiveness and practicability of the proposed We-Market,which achieves 15%cost reduction with shorter running time.Comparative analysis indicates better scalability,which is more suitable for largerscale We-Market implementation.展开更多
With the development of the Internet of Things and devices continuing to scale,using cloud computing resources to process data in real-time is challenging.Edge computing technologies can improve real-time performance ...With the development of the Internet of Things and devices continuing to scale,using cloud computing resources to process data in real-time is challenging.Edge computing technologies can improve real-time performance in processing data.By introducing the FPGA into the computing node and using the dynamic reconfigurability of the FPGA,the FPGA-based edge node can increase the edge node capability.In this paper,a task-based collaborative method for an FPGA-based edge computing system is proposed in order to meet the collaboration among FPGA-based edge nodes,edge nodes,and the cloud.The modeling of the task includes two parts,task information and task-dependent file.Task information is used to describe the running information and dependency infor-mation required for the task execution.Task-dependent file contains the configuration bit-stream of FPGA in running of the task.By analyzing the task behavior,this paper builds four basic behaviors,analyzes the critical attributes of each behavior,and summa-rizes the task model suitable for FPGA-based edge nodes.Tasks with specific functions can be created by modifying different attributes of model nodes.Finally,the availability of the model and the task-based collaborative method are verified by simulation exper-iments.The experimental results that the task model proposed in this paper can meet cloud-edge collaboration in the FPGA-based edge computing environment.展开更多
The deep neural network is a reliable technical support for cloud com-puting and edge computing.It has excellent nonlinear approximation and gener-alization capabilities,making it suitable for classifying and predicti...The deep neural network is a reliable technical support for cloud com-puting and edge computing.It has excellent nonlinear approximation and gener-alization capabilities,making it suitable for classifying and predicting Internet of Things data in cloud computing and edge computingfields.However,the increas-ing size of neural networks poses a challenge for their deployment on devices with limited computing and storage resources.Traditional cloud computing ser-vices also suffer from high latency,which hinders real-time tasks.To address these challenges,this paper proposes a cloud-side cooperation model for deep learning based on migration learning technology.This model used migration learning tech-nology to reduce the size of deep neural networks.Specifically,it deployed the deep neural network model(CDLM)in the cloud and the shallow neural network model(EDLM)at the edge.CDLM is used to help train EDLM and improve its performance,enabling it to run independently on edge devices with high accu-racy and respond to real-time tasks.This approach reduced the amount of user data transmitted to the cloud,alleviated bandwidth pressure,and protected user privacy.Experimental results show that the proposed model improved the accu-racy of EDLM by 19.58% compared with traditional neural network models.Thesefindings provide a theoretical and experimental foundation for the study of cloud-edge collaborative models.展开更多
The telecommunications industry is becoming increasingly aware of potential subscriber churn as a result of the growing popularity of smartphones in the mobile Internet era,the quick development of telecommunications ...The telecommunications industry is becoming increasingly aware of potential subscriber churn as a result of the growing popularity of smartphones in the mobile Internet era,the quick development of telecommunications services,the implementation of the number portability policy,and the intensifying competition among operators.At the same time,users'consumption preferences and choices are evolving.Excellent churn prediction models must be created in order to accurately predict the churn tendency,since keeping existing customers is far less expensive than acquiring new ones.But conventional or learning-based algorithms can only go so far into a single subscriber's data;they cannot take into consideration changes in a subscriber's subscription and ignore the coupling and correlation between various features.Additionally,the current churn prediction models have a high computational burden,a fuzzy weight distribution,and significant resource economic costs.The prediction algorithms involving network models currently in use primarily take into account the private information shared between users with text and pictures,ignoring the reference value supplied by other users with the same package.This work suggests a user churn prediction model based on Graph Attention Convolutional Neural Network(GAT-CNN)to address the aforementioned issues.The main contributions of this paper are as follows:Firstly,we present a three-tiered hierarchical cloud-edge cooperative framework that increases the volume of user feature input by means of two aggregations at the device,edge,and cloud layers.Second,we extend the use of users'own data by introducing self-attention and graph convolution models to track the relative changes of both users and packages simultaneously.Lastly,we build an integrated offline-online system for churn prediction based on the strengths of the two models,and we experimentally validate the efficacy of cloudside collaborative training and inference.In summary,the churn prediction model based on Graph Attention Convolutional Neural Network presented in this paper can effectively address the drawbacks of conventional algorithms and offer telecom operators crucial decision support in developing subscriber retention strategies and cutting operational expenses.展开更多
The increasing integration of intermittent renewable energy sources into distribution networks has exerted significant pressure on the frequency regulation of power systems.Meanwhile,integrating small-capacity battery...The increasing integration of intermittent renewable energy sources into distribution networks has exerted significant pressure on the frequency regulation of power systems.Meanwhile,integrating small-capacity battery energy storage systems into distribution network is a growing trend in the construction of virtual power plants(VPPs),which offer great potential advantages in improving the system frequency regulation capabilities.However,the process of power dispatch for VPPs may be hindered by imperfections in the communication network,which affects their frequency control performance.Simultaneously,the economic benefits associated with their frequency control services are often overlooked.As such,we propose a codesign method of power dispatch with dynamic power regulation and communication transmission optimization for frequency control in VPPs.First,a joint design scheme of power dispatch and routing optimization under cloud-edge collaborations is proposed.This scheme encompasses a power dispatch method considering the influences of communication network and a routing optimization policy based on graph convolutional neural networks,both of which are designed to ensure the accurate and real-time frequency control service.Further,we propose a dynamic power regulation strategy under edge-edge collaborations.Specifically,according to the established correction control objective,an adaptive distributed auction algorithm(ADAA)based dynamic power regulation control method is designed to determine the optimal regulation power of VPPs,thereby improving the economic benefits of frequency control service.Finally,the simulation results validate the feasibility and superiority of the proposed co-design method for frequency control.展开更多
The study aims to address the challenge of dynamic assessment in power systems by proposing a design scheme for an intelligent adaptive power distribution system based on runtime verification.The system architecture i...The study aims to address the challenge of dynamic assessment in power systems by proposing a design scheme for an intelligent adaptive power distribution system based on runtime verification.The system architecture is built upon cloud-edge-end collaboration,enabling comprehensive monitoring and precise management of the power grid through coordinated efforts across different levels.Specif-ically,the study employs the adaptive observer approach,allowing dynamic adjustments to observers to reflect updates in requirements and ensure system reliability.This method covers both structural and parametric adjustments to specifications,including updating time protection conditions,updating events,and adding or removing responses.The results demonstrate that with the implementation of adaptive observers,the system becomes more flexible in responding to changes,significantly enhancing its level of efficiency.By employing dynamically changing verification specifications,the system achieves real-time and flexible verification.This research provides technical support for the safe,efficient,and reliable operation of electrical power distribution systems.展开更多
As the application of Industrial Robots(IRs)scales and related participants increase,the demands for intelligent Operation and Maintenance(O&M)and multi-tenant collaboration rise.Traditional methods could no longe...As the application of Industrial Robots(IRs)scales and related participants increase,the demands for intelligent Operation and Maintenance(O&M)and multi-tenant collaboration rise.Traditional methods could no longer cover the requirements,while the Industrial Internet of Things(IIoT)has been considered a promising solution.However,there’s a lack of IIoT platforms dedicated to IR O&M,including IR maintenance,process optimization,and knowledge sharing.In this context,this paper puts forward the multi-tenant-oriented ACbot platform,which attempts to provide the first holistic IIoT-based solution for O&M of IRs.Based on an information model designed for the IR field,ACbot has implemented an application architecture with resource and microservice management across the cloud and multiple edges.On this basis,we develop four vital applications including real-time monitoring,health management,process optimization,and knowledge graph.We have deployed the ACbot platform in real-world scenarios that contain various participants,types of IRs,and processes.To date,ACbot has been accessed by 10 organizations and managed 60 industrial robots,demonstrating that the platform fulfills our expectations.Furthermore,the application results also showcase its robustness,versatility,and adaptability for developing and hosting intelligent robot applications.展开更多
Vehicle Edge Computing(VEC)and Cloud Computing(CC)significantly enhance the processing efficiency of delay-sensitive and computation-intensive applications by offloading compute-intensive tasks from resource-constrain...Vehicle Edge Computing(VEC)and Cloud Computing(CC)significantly enhance the processing efficiency of delay-sensitive and computation-intensive applications by offloading compute-intensive tasks from resource-constrained onboard devices to nearby Roadside Unit(RSU),thereby achieving lower delay and energy consumption.However,due to the limited storage capacity and energy budget of RSUs,it is challenging to meet the demands of the highly dynamic Internet of Vehicles(IoV)environment.Therefore,determining reasonable service caching and computation offloading strategies is crucial.To address this,this paper proposes a joint service caching scheme for cloud-edge collaborative IoV computation offloading.By modeling the dynamic optimization problem using Markov Decision Processes(MDP),the scheme jointly optimizes task delay,energy consumption,load balancing,and privacy entropy to achieve better quality of service.Additionally,a dynamic adaptive multi-objective deep reinforcement learning algorithm is proposed.Each Double Deep Q-Network(DDQN)agent obtains rewards for different objectives based on distinct reward functions and dynamically updates the objective weights by learning the value changes between objectives using Radial Basis Function Networks(RBFN),thereby efficiently approximating the Pareto-optimal decisions for multiple objectives.Extensive experiments demonstrate that the proposed algorithm can better coordinate the three-tier computing resources of cloud,edge,and vehicles.Compared to existing algorithms,the proposed method reduces task delay and energy consumption by 10.64%and 5.1%,respectively.展开更多
文摘Using satellites to complete spectrum monitoring tasks can effectively receive and process electromagnetic spectrum signals emitted by radiation sources.However,due to the shortage of satellite storage,computing and network resources,the intersatellite coordination is weak,and with the massive growth of spectrum data,the traditional cloud computing mode cannot meet the requirements of electromagnetic spectrum monitoring in terms of real-time,bandwidth,and security.We apply edge computing technology and deep learning technology to the satellite.Aiming at the problems of distributed satellite management and control,we propose a space-based distributed electromagnetic spectrum monitoring intelligent connected cloud-edge collaborative architecture SpaceEdge.SpaceEdge applies edge computing and artificial intelligence technology to space-based spectrum monitoring.SpaceEdge deploys intelligent monitoring algorithms to edge nodes to form edge intelligent satellite,and uses the cloud to uniformly manage and control heterogeneous edge satellite and monitor satellite resources.In addition,SpaceEdge can also adjust edge intelligent spectrum monitoring applications as needed to achieve effective coordination of inter-satellite algorithms and data to achieve the purpose of collaborative monitoring.Finally,SpaceEdge was experimentally verified,and the results proved the feasibility of SpaceEdge and can improve the timeliness and autonomy of the distributed satellite’s coordinated signal monitoring.
基金This work was supported by the Key Research and Development(R&D)Plan of Heilongjiang Province of China(JD22A001).
文摘With the continuous development of network func-tions virtualization(NFV)and software-defined networking(SDN)technologies and the explosive growth of network traffic,the requirement for computing resources in the network has risen sharply.Due to the high cost of edge computing resources,coordinating the cloud and edge computing resources to improve the utilization efficiency of edge computing resources is still a considerable challenge.In this paper,we focus on optimiz-ing the placement of network services in cloud-edge environ-ments to maximize the efficiency.It is first proved that,in cloud-edge environments,placing one service function chain(SFC)integrally in the cloud or at the edge can improve the utilization efficiency of edge resources.Then a virtual network function(VNF)performance-resource(P-R)function is proposed to repre-sent the relationship between the VNF instance computing per-formance and the allocated computing resource.To select the SFCs that are most suitable to deploy at the edge,a VNF place-ment and resource allocation model is built to configure each VNF with its particular P-R function.Moreover,a heuristic recur-sive algorithm is designed called the recursive algorithm for max edge throughput(RMET)to solve the model.Through simula-tions on two scenarios,it is verified that RMET can improve the utilization efficiency of edge computing resources.
基金supported by the China Postdoctoral Science Foundation (Grant Nos. 2021TQ0042, 2021M700435, 2021TQ0041)the National Natural Science Foundation of China (Grant No. 62102027)the Shandong Provincial Key Research and Development Program (2021CXGC010106)
文摘In this paper,we propose a novel fuzzy matching data sharing scheme named FADS for cloudedge communications.FADS allows users to specify their access policies,and enables receivers to obtain the data transmitted by the senders if and only if the two sides meet their defined certain policies simultaneously.Specifically,we first formalize the definition and security models of fuzzy matching data sharing in cloud-edge environments.Then,we construct a concrete instantiation by pairing-based cryptosystem and the privacy-preserving set intersection on attribute sets from both sides to construct a concurrent matching over the policies.If the matching succeeds,the data can be decrypted.Otherwise,nothing will be revealed.In addition,FADS allows users to dynamically specify the policy for each time,which is an urgent demand in practice.A thorough security analysis demonstrates that FADS is of provable security under indistinguishable chosen ciphertext attack(IND-CCA)in random oracle model against probabilistic polynomial-time(PPT)adversary,and the desirable security properties of privacy and authenticity are achieved.Extensive experiments provide evidence that FADS is with acceptable efficiency.
基金supported by Key-Area Research and Development Program of Guangdong Province(2021B0101420002)the Major Key Project of PCL(PCL2021A09)+3 种基金National Natural Science Foundation of China(62072187)Guangdong Major Project of Basic and Applied Basic Research(2019B030302002)Guangdong Marine Economic Development Special Fund Project(GDNRC[2022]17)Guangzhou Development Zone Science and Technology(2021GH10,2020GH10).
文摘Nowadays,the rapid development of edge computing has driven an increasing number of deep learning applications deployed at the edge of the network,such as pedestrian and vehicle detection,to provide efficient intelligent services to mobile users.However,as the accuracy requirements continue to increase,the components of deep learning models for pedestrian and vehicle detection,such as YOLOv4,become more sophisticated and the computing resources required for model training are increasing dramatically,which in turn leads to significant challenges in achieving effective deployment on resource-constrained edge devices while ensuring the high accuracy performance.For addressing this challenge,a cloud-edge collaboration-based pedestrian and vehicle detection framework is proposed in this paper,which enables sufficient training of models by utilizing the abundant computing resources in the cloud,and then deploying the well-trained models on edge devices,thus reducing the computing resource requirements for model training on edge devices.Furthermore,to reduce the size of the model deployed on edge devices,an automatic pruning method combines the convolution layer and BN layer is proposed to compress the pedestrian and vehicle detection model size.Experimental results show that the framework proposed in this paper is able to deploy the pruned model on a real edge device,Jetson TX2,with 6.72 times higher FPS.Meanwhile,the channel pruning reduces the volume and the number of parameters to 96.77%for the model,and the computing amount is reduced to 81.37%.
基金supported by the National Natural Science Foundation of China under Grant 52077146.
文摘With the construction of the power Internet of Things(IoT),communication between smart devices in urban distribution networks has been gradually moving towards high speed,high compatibility,and low latency,which provides reliable support for reconfiguration optimization in urban distribution networks.Thus,this study proposed a deep reinforcement learning based multi-level dynamic reconfiguration method for urban distribution networks in a cloud-edge collaboration architecture to obtain a real-time optimal multi-level dynamic reconfiguration solution.First,the multi-level dynamic reconfiguration method was discussed,which included feeder-,transformer-,and substation-levels.Subsequently,the multi-agent system was combined with the cloud-edge collaboration architecture to build a deep reinforcement learning model for multi-level dynamic reconfiguration in an urban distribution network.The cloud-edge collaboration architecture can effectively support the multi-agent system to conduct“centralized training and decentralized execution”operation modes and improve the learning efficiency of the model.Thereafter,for a multi-agent system,this study adopted a combination of offline and online learning to endow the model with the ability to realize automatic optimization and updation of the strategy.In the offline learning phase,a Q-learning-based multi-agent conservative Q-learning(MACQL)algorithm was proposed to stabilize the learning results and reduce the risk of the next online learning phase.In the online learning phase,a multi-agent deep deterministic policy gradient(MADDPG)algorithm based on policy gradients was proposed to explore the action space and update the experience pool.Finally,the effectiveness of the proposed method was verified through a simulation analysis of a real-world 445-node system.
基金supported in part by the National Natural Science Foundation of China under Grant 62162018 and Grant 61861013in part by the Innovation Research Team Project of Guangxi Natural Science Foundation 2019GXNSFGA245004.
文摘Software-defined networking(SDN)enables the separation of control and data planes,allowing for centralized control and management of the network.Without adequate access control methods,the risk of unau-thorized access to the network and its resources increases significantly.This can result in various security breaches.In addition,if authorized devices are attacked or controlled by hackers,they may turn into malicious devices,which can cause severe damage to the network if their abnormal behaviour goes undetected and their access privileges are not promptly restricted.To solve those problems,an anomaly detection and access control mechanism based on SDN and neural networks is proposed for cloud-edge collaboration networks.The system employs the Attribute Based Access Control(ABAC)model and smart contract for fine-grained control of device access to the network.Furthermore,a cloud-edge collaborative Key Performance Indicator(KPI)anomaly detection method based on the Gated Recurrent Unit and Generative Adversarial Nets(GRU-GAN)is designed to discover the anomaly devices.An access restriction mechanism based on reputation value and anomaly detection is given to prevent anomalous devices.Experiments show that the proposed mechanism performs better anomaly detection on several datasets.The reputation-based access restriction effectively reduces the number of malicious device attacks.
基金supported by the National Natural Science Foundation of China(Nos.62162018,61972412)the Natural Science Foundation of Guangxi(No.2019GXNSFGA245004)+1 种基金the Guilin Science and Technology Project(20210226-1)the Innovation Project of Guangxi Graduate Education(No.YCSW2022296).
文摘Cloud storage and edge computing are utilized to address the storage and computational challenges arising from the exponential data growth in IoT.However,data privacy is potentially risky when data is outsourced to cloud servers or edge services.While data encryption ensures data confidentiality,it can impede data sharing and retrieval.Attribute-based searchable encryption(ABSE)is proposed as an effective technique for enhancing data security and privacy.Nevertheless,ABSE has its limitations,such as single attribute authorization failure,privacy leakage during the search process,and high decryption overhead.This paper presents a novel approach called the blockchain-assisted efficientmulti-authority attribute-based searchable encryption scheme(BEM-ABSE)for cloudedge collaboration scenarios to address these issues.BEM-ABSE leverages a consortium blockchain to replace the central authentication center for global public parameter management.It incorporates smart contracts to facilitate reliable and fair ciphertext keyword search and decryption result verification.To minimize the computing burden on resource-constrained devices,BEM-ABSE adopts an online/offline hybrid mechanism during the encryption process and a verifiable edge-assisted decryption mechanism.This ensures both low computation cost and reliable ciphertext.Security analysis conducted under the random oracle model demonstrates that BEM-ABSE is resistant to indistinguishable chosen keyword attacks(IND-CKA)and indistinguishable chosen plaintext attacks(INDCPA).Theoretical analysis and simulation results confirm that BEM-ABSE significantly improves computational efficiency compared to existing solutions.
文摘As the penetration rate of renewable energy sources(RES)gradually increases,demand-side resources(DSR)should be fully utilized to provide flexibility and rapidly respond to real-time power supply-demand imbalance.However,scheduling a large number of DSR clusters will inevitably bring unbearable transmission delay,and computation delay,which in turn lead to lower response speeds.This paper examines flexibility scheduling of DSR clusters within a smart distribution network(SDN)in view of both kinds of delay.Building upon a SDN model,maximum schedulable flexibility of DSR clusters is first quantified.Then,a flexibility response curve is analyzed to reflect the effect of delay on flexibility scheduling.Aiming at reducing flexibility shortage brought by delay,we propose a modified flexibility scheduling strategy based on cloud-edge collaboration.Compared with traditional strategy,centralized optimization is replaced by distributed optimization to consider both economic efficiency and effect of delay.Besides,an offloading strategy is also formulated to decide optimal edge nodes and corresponding wired paths for edge computations.In a case study,we evaluate scheduled flexibility,operational cost,average delay and the chosen edge nodes for edge computations with traditional strategy and our proposed strategy.Evaluation results show the proposed strategy can significantly reduce the effect of delay on flexibility scheduling,and guarantee the optimality of operational cost to some extent.
基金supported by the Natural Science Foundation of Zhejiang Province(Nos.LQ21F020021 and LZ21F020008)Zhejiang Provincial Natural Science Foundation of China(No.LZ22F020002)the Research Start-up Project funded by Hangzhou Normal University(No.2020QD2035).
文摘Service recommendation provides an effective solution to extract valuable information from the huge and ever-increasing volume of big data generated by the large cardinality of user devices.However,the distributed and rich multi-source big data resources raise challenges to the centralized cloud-based data storage and value mining approaches in terms of economic cost and effective service recommendation methods.In view of these challenges,we propose a deep neural collaborative filtering based service recommendation method with multi-source data(i.e.,NCF-MS)in this paper,which adopts the cloud-edge collaboration computing paradigm to build recommendation model.More specifically,the Stacked Denoising Auto Encoder(SDAE)module is adopted to extract user/service features from auxiliary user profiles and service attributes.The Multiple Layer Perceptron(MLP)module is adopted to integrate the auxiliary user/service features to train the recommendation model.Finally,we evaluate the effectiveness of the NCF-MS method on three public datasets.The experimental results show that our proposed method achieves better performance than existing methods.
基金supported by the National Natural Science Foundation of China(Nos.62273292,62276226,and 61973261)。
文摘How to collaboratively offload tasks between user devices,edge networks(ENs),and cloud data centers is an interesting and challenging research topic.In this paper,we investigate the offoading decision,analytical modeling,and system parameter optimization problem in a collaborative cloud-edge device environment,aiming to trade off different performance measures.According to the differentiated delay requirements of tasks,we classify the tasks into delay-sensitive and delay-tolerant tasks.To meet the delay requirements of delay-sensitive tasks and process as many delay-tolerant tasks as possible,we propose a cloud-edge device collaborative task offoading scheme,in which delay-sensitive and delay-tolerant tasks follow the access threshold policy and the loss policy,respectively.We establish a four-dimensional continuous-time Markov chain as the system model.By using the Gauss-Seidel method,we derive the stationary probability distribution of the system model.Accordingly,we present the blocking rate of delay-sensitive tasks and the average delay of these two types of tasks.Numerical experiments are conducted and analyzed to evaluate the system performance,and numerical simulations are presented to evaluate and validate the effectiveness of the proposed task offloading scheme.Finally,we optimize the access threshold in the EN buffer to obtain the minimum system cost with different proportions of delay-sensitive tasks.
基金supported by the National Natural Science Foundation of China(No.U1966204,No.52122706)。
文摘Market participants can only bid with lagged information disclosure under the existing market mechanism,which can lead to information asymmetry and irrational market behavior,thus influencing market efficiency.To promote rational bidding behavior of market participants and improve market efficiency,a novel electricity market mechanism based on cloudedge collaboration is proposed in this paper.Critical market information,called residual demand curve,is published to market participants in real-time on the cloud side,while participants on the edge side are allowed to adjust their bids according to the information disclosure prior to closure gate.The proposed mechanism can encourage rational bids in an incentive-compatible way through the process of dynamic equilibrium while protecting participants’privacy.This paper further formulates the mathematical model of market equilibrium to simulate the process of each market participant’s strategic bidding behavior towards equilibrium.A case study based on the IEEE 30-bus system shows the proposed market mechanism can effectively guide bidding behavior of market participants,while condensing exchanged information and protecting privacy of participants.
基金This work was supported by the Science and Technology Program of State Grid Corporation of China(No.521002190049).
文摘With the increasing penetration of renewable energy generation,uncertainty and randomness pose great challenges for optimal dispatching in distribution networks.We propose a cloud-edge cooperative dispatching(CECD)method to exploit the new opportunities offered by Internet of Things(IoT)technology.To alleviate the huge pressure on the modeling and computing of large-scale distribution system,the method deploys edge nodes in small-scale transformer areas in which robust optimization subproblem models are introduced to address the photovoltaic(PV)uncertainty.Considering the limited communication and computing capabilities of the edge nodes,the cloud center in the distribution automation system(DAS)establishes a utility grid master problem model that enforces the consistency between the solution at each edge node with the utility grid based on the alternating direction method of multipliers(ADMM).Furthermore,the voltage constraint derived from the linear power flow equations is adopted for enhancing the operation security of the distribution network.We perform a cloud-edge system simulation of the proposed CECD method and demonstrate a dispatching application.The case study is carried out on a modified 33-node system to verify the remarkable performance of the proposed model and method.
基金supported in part by the National Natural Science Foundation of China(No.U1908213)Colleges and Universities in Hebei Province Science Research Program(No.QN2020504)Fundamental Research Funds for the Central Universities(No.N2223001)。
文摘With the extensive penetration of distributed renewable energy and self-interested prosumers,the emerging power market tends to enable user autonomy by bottom-up control and distributed coordination.This paper is devoted to solving the specific problems of distributed energy management and autonomous bidding and peer-to-peer(P2P)energy sharing among prosumers.A novel cloud-edge-based We-Market is presented,where the prosumers,as edge nodes with independent control,balance the electricity cost and thermal comfort by formulating a dynamic household energy management system(HEMS).Meanwhile,the autonomous bidding is initiated by prosumers via the modified Stone-Geary utility function.In the cloud center,a distributed convergence bidding(CB)algorithm based on consistency criterion is developed,which promotes faster and fairer bidding through the interactive iteration with the edge nodes.Besides,the proposed scheme is built on top of the commercial cloud platform with sufficiently secure and scalable computing capacity.Numerical results show the effectiveness and practicability of the proposed We-Market,which achieves 15%cost reduction with shorter running time.Comparative analysis indicates better scalability,which is more suitable for largerscale We-Market implementation.
基金This work is supported by the National Key R&D Program of China(Grant No.2018YFB1701600).
文摘With the development of the Internet of Things and devices continuing to scale,using cloud computing resources to process data in real-time is challenging.Edge computing technologies can improve real-time performance in processing data.By introducing the FPGA into the computing node and using the dynamic reconfigurability of the FPGA,the FPGA-based edge node can increase the edge node capability.In this paper,a task-based collaborative method for an FPGA-based edge computing system is proposed in order to meet the collaboration among FPGA-based edge nodes,edge nodes,and the cloud.The modeling of the task includes two parts,task information and task-dependent file.Task information is used to describe the running information and dependency infor-mation required for the task execution.Task-dependent file contains the configuration bit-stream of FPGA in running of the task.By analyzing the task behavior,this paper builds four basic behaviors,analyzes the critical attributes of each behavior,and summa-rizes the task model suitable for FPGA-based edge nodes.Tasks with specific functions can be created by modifying different attributes of model nodes.Finally,the availability of the model and the task-based collaborative method are verified by simulation exper-iments.The experimental results that the task model proposed in this paper can meet cloud-edge collaboration in the FPGA-based edge computing environment.
基金This work is supported by the following projects:Natural Science Foundation of Jilin Province of China(Grant No.20220101136JC).
文摘The deep neural network is a reliable technical support for cloud com-puting and edge computing.It has excellent nonlinear approximation and gener-alization capabilities,making it suitable for classifying and predicting Internet of Things data in cloud computing and edge computingfields.However,the increas-ing size of neural networks poses a challenge for their deployment on devices with limited computing and storage resources.Traditional cloud computing ser-vices also suffer from high latency,which hinders real-time tasks.To address these challenges,this paper proposes a cloud-side cooperation model for deep learning based on migration learning technology.This model used migration learning tech-nology to reduce the size of deep neural networks.Specifically,it deployed the deep neural network model(CDLM)in the cloud and the shallow neural network model(EDLM)at the edge.CDLM is used to help train EDLM and improve its performance,enabling it to run independently on edge devices with high accu-racy and respond to real-time tasks.This approach reduced the amount of user data transmitted to the cloud,alleviated bandwidth pressure,and protected user privacy.Experimental results show that the proposed model improved the accu-racy of EDLM by 19.58% compared with traditional neural network models.Thesefindings provide a theoretical and experimental foundation for the study of cloud-edge collaborative models.
基金supported by National Key R&D Program of China(No.2022YFB3104500)Natural Science Foundation of Jiangsu Province(No.BK20222013)Scientific Research Foundation of Nanjing Institute of Technology(No.3534113223036)。
文摘The telecommunications industry is becoming increasingly aware of potential subscriber churn as a result of the growing popularity of smartphones in the mobile Internet era,the quick development of telecommunications services,the implementation of the number portability policy,and the intensifying competition among operators.At the same time,users'consumption preferences and choices are evolving.Excellent churn prediction models must be created in order to accurately predict the churn tendency,since keeping existing customers is far less expensive than acquiring new ones.But conventional or learning-based algorithms can only go so far into a single subscriber's data;they cannot take into consideration changes in a subscriber's subscription and ignore the coupling and correlation between various features.Additionally,the current churn prediction models have a high computational burden,a fuzzy weight distribution,and significant resource economic costs.The prediction algorithms involving network models currently in use primarily take into account the private information shared between users with text and pictures,ignoring the reference value supplied by other users with the same package.This work suggests a user churn prediction model based on Graph Attention Convolutional Neural Network(GAT-CNN)to address the aforementioned issues.The main contributions of this paper are as follows:Firstly,we present a three-tiered hierarchical cloud-edge cooperative framework that increases the volume of user feature input by means of two aggregations at the device,edge,and cloud layers.Second,we extend the use of users'own data by introducing self-attention and graph convolution models to track the relative changes of both users and packages simultaneously.Lastly,we build an integrated offline-online system for churn prediction based on the strengths of the two models,and we experimentally validate the efficacy of cloudside collaborative training and inference.In summary,the churn prediction model based on Graph Attention Convolutional Neural Network presented in this paper can effectively address the drawbacks of conventional algorithms and offer telecom operators crucial decision support in developing subscriber retention strategies and cutting operational expenses.
基金supported in part by the Major Program of the National Natural Science Foundation of China(No.62293504)the Postgraduate Research&Practice Innovation Program of Jiangsu Province(No.KYCX24_1212)。
文摘The increasing integration of intermittent renewable energy sources into distribution networks has exerted significant pressure on the frequency regulation of power systems.Meanwhile,integrating small-capacity battery energy storage systems into distribution network is a growing trend in the construction of virtual power plants(VPPs),which offer great potential advantages in improving the system frequency regulation capabilities.However,the process of power dispatch for VPPs may be hindered by imperfections in the communication network,which affects their frequency control performance.Simultaneously,the economic benefits associated with their frequency control services are often overlooked.As such,we propose a codesign method of power dispatch with dynamic power regulation and communication transmission optimization for frequency control in VPPs.First,a joint design scheme of power dispatch and routing optimization under cloud-edge collaborations is proposed.This scheme encompasses a power dispatch method considering the influences of communication network and a routing optimization policy based on graph convolutional neural networks,both of which are designed to ensure the accurate and real-time frequency control service.Further,we propose a dynamic power regulation strategy under edge-edge collaborations.Specifically,according to the established correction control objective,an adaptive distributed auction algorithm(ADAA)based dynamic power regulation control method is designed to determine the optimal regulation power of VPPs,thereby improving the economic benefits of frequency control service.Finally,the simulation results validate the feasibility and superiority of the proposed co-design method for frequency control.
基金supported by the China Electric Power ResearchInstitute and Electric Power Research Institute State Grid AnhuiElectric Power Co.,Ltd.,China(5400-202355201A-1-1-ZN).
文摘The study aims to address the challenge of dynamic assessment in power systems by proposing a design scheme for an intelligent adaptive power distribution system based on runtime verification.The system architecture is built upon cloud-edge-end collaboration,enabling comprehensive monitoring and precise management of the power grid through coordinated efforts across different levels.Specif-ically,the study employs the adaptive observer approach,allowing dynamic adjustments to observers to reflect updates in requirements and ensure system reliability.This method covers both structural and parametric adjustments to specifications,including updating time protection conditions,updating events,and adding or removing responses.The results demonstrate that with the implementation of adaptive observers,the system becomes more flexible in responding to changes,significantly enhancing its level of efficiency.By employing dynamically changing verification specifications,the system achieves real-time and flexible verification.This research provides technical support for the safe,efficient,and reliable operation of electrical power distribution systems.
基金supported by the Zhejiang Province Key R&D Program of China(2023C01070).
文摘As the application of Industrial Robots(IRs)scales and related participants increase,the demands for intelligent Operation and Maintenance(O&M)and multi-tenant collaboration rise.Traditional methods could no longer cover the requirements,while the Industrial Internet of Things(IIoT)has been considered a promising solution.However,there’s a lack of IIoT platforms dedicated to IR O&M,including IR maintenance,process optimization,and knowledge sharing.In this context,this paper puts forward the multi-tenant-oriented ACbot platform,which attempts to provide the first holistic IIoT-based solution for O&M of IRs.Based on an information model designed for the IR field,ACbot has implemented an application architecture with resource and microservice management across the cloud and multiple edges.On this basis,we develop four vital applications including real-time monitoring,health management,process optimization,and knowledge graph.We have deployed the ACbot platform in real-world scenarios that contain various participants,types of IRs,and processes.To date,ACbot has been accessed by 10 organizations and managed 60 industrial robots,demonstrating that the platform fulfills our expectations.Furthermore,the application results also showcase its robustness,versatility,and adaptability for developing and hosting intelligent robot applications.
基金supported by Key Science and Technology Program of Henan Province,China(Grant Nos.242102210147,242102210027)Fujian Province Young and Middle aged Teacher Education Research Project(Science and Technology Category)(No.JZ240101)(Corresponding author:Dong Yuan).
文摘Vehicle Edge Computing(VEC)and Cloud Computing(CC)significantly enhance the processing efficiency of delay-sensitive and computation-intensive applications by offloading compute-intensive tasks from resource-constrained onboard devices to nearby Roadside Unit(RSU),thereby achieving lower delay and energy consumption.However,due to the limited storage capacity and energy budget of RSUs,it is challenging to meet the demands of the highly dynamic Internet of Vehicles(IoV)environment.Therefore,determining reasonable service caching and computation offloading strategies is crucial.To address this,this paper proposes a joint service caching scheme for cloud-edge collaborative IoV computation offloading.By modeling the dynamic optimization problem using Markov Decision Processes(MDP),the scheme jointly optimizes task delay,energy consumption,load balancing,and privacy entropy to achieve better quality of service.Additionally,a dynamic adaptive multi-objective deep reinforcement learning algorithm is proposed.Each Double Deep Q-Network(DDQN)agent obtains rewards for different objectives based on distinct reward functions and dynamically updates the objective weights by learning the value changes between objectives using Radial Basis Function Networks(RBFN),thereby efficiently approximating the Pareto-optimal decisions for multiple objectives.Extensive experiments demonstrate that the proposed algorithm can better coordinate the three-tier computing resources of cloud,edge,and vehicles.Compared to existing algorithms,the proposed method reduces task delay and energy consumption by 10.64%and 5.1%,respectively.
