检索结果-内蒙古大学图书馆

Asynchronous Federated Deep Reinforcement Learning-Based URLLC-Aware computation offloading in Space-Assisted Vehicular Networks

引用

IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS 2023年第7期24卷 7377-7389页

作者： Pan, Chao Wang, Zhao Liao, Haijun Zhou, Zhenyu Wang, Xiaoyan Tariq, Muhammad Al-Otaibi, Sattam North China Elect Power Univ State Key Lab Alternate Elect Power Syst Renewabl Beijing 102206 Peoples R China Southeast Univ Natl Mobile Commun Res Lab Nanjing 211189 Peoples R China Ibaraki Univ Grad Sch Sci & Engn Ibaraki 3108512 Japan Natl Univ Comp & Emerging Sci Elect Engn Dept Islamabad 44000 Pakistan Taif Univ Coll Engn Innovat & Entrepreneurship Ctr At Taif 26571 Saudi Arabia

Space-assisted vehicular networks (SAVN) provide seamless coverage and on-demand data processing services for user vehicles (UVs). However, ultra-reliable and low-latency communication (URLLC) demands imposed by emerging vehicular applications are hard to be satisfied in SAVN by existing computation offloading techniques. Traditional deep reinforcement learning algorithms are unsuitable for highly dynamic SAVN due to the underutilization of environment observations. An AsynchronouS federaTed deep Q-learning (DQN)-basEd and URLLC-aware computation offloading algorithm (ASTEROID) is presented in this paper to achieve throughput maximization considering the long-term URLLC constraints. Specifically, we first establish an extreme value theory-based URLLC constraint model. Second, the task offloading and computation resource allocation are decomposed by employing Lyapunov optimization. Finally, an asynchronous federated DQN-based (AF-DQN) algorithm is presented to address the UV-side task offloading problem. The server-side computation resource allocation is settled by an queue backlog-aware algorithm. Simulation results verify that ASTEROID achieves superior throughput and URLLC performances.

关键词： Ultra reliable low latency communication Task analysis computational modeling Servers Optimization Heuristic algorithms Throughput Space-assisted vehicular networks (SAVN) ultra-reliable and low-latency communication awareness asynchronous federated deep reinforcement learning computation offloading

来源：评论

学校读者我要写书评

暂无评论

Deep reinforcement learning based computation offloading and resource allocation strategy for maritime internet of things

引用

COMPUTER NETWORKS 2025年 264卷

作者： Xu, Yanli Yu, Qianlian Shanghai Maritime Univ Sch Informat Engn Shanghai 201306 Peoples R China

With the continuous growth of marine activities, the demands of computation-intensive applications dramatically increase. How to provide guaranteed computation service for maritime Internet of Things (M-IoT) is not well addressed. In this paper, we make use of space-air-ground-sea integrated network (SAGSIN) and mobile edge computing (MEC) to provide wide-coverage and efficient computation services for M-IoT users. The heterogeneous network resources of SAGSIN are fully used by unified management based on software-defined networking (SDN). In addition, we propose a joint computation offloading and resource allocation algorithm based on dueling double deep Q network (D3QN) with the consideration of heterogeneous SAGSIN architecture and scarce service resources of marine scenarios. The proposed algorithm can minimize delays and energy consumption by optimizing task offloading decisions and resource allocation of communication and computation. It achieves efficient resource utilization and ensures that the processing power is highly matched with the task requirements. The simulation results demonstrate that the proposed algorithm outperforms Deep Q Network (DQN), Double DQN, and Dueling DQN in terms of the overall system overhead. Furthermore, the proposed algorithm exhibits exceptional performance and robustness. It can ensure the increasing demand for computing services in Marine networks.

关键词： Maritime Internet of Things (M-IoT) Mobile edge computing (MEC) computation offloading Resource allocation Deep reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Hierarchical Deep Reinforcement Learning for Joint Service Caching and computation offloading in Mobile Edge-Cloud Computing

引用

IEEE TRANSACTIONS ON SERVICES COMPUTING 2024年第4期17卷 1548-1564页

作者： Sun, Chuan Li, Xiuhua Wang, Chenyang He, Qiang Wang, Xiaofei Leung, Victor C. M. Chongqing Univ Sch Big Data & Software Engn Chongqing 400044 Peoples R China Tianjin Univ Coll Intelligence & Comp Tianjin 301600 Peoples R China Huazhong Univ Sci & Technol Serv Comp Technol & Syst Lab Natl Engn Res Ctr Big Data Technol & Syst Sch Comp Sci & TechnolCluster & Grid Comp Lab Wuhan 430074 Peoples R China Shenzhen Univ Coll Comp Sci & Software Engn Shenzhen 518060 Peoples R China Univ British Columbia Dept Elect & Comp Engn Vancouver BC V6T1Z4 Canada

Mobile edge-cloud computing networks can provide distributed, hierarchical, and fine-grained resources, and have become a major goal for future high-performance computing networks. The key is how to jointly optimize service caching and computation offloading. However, the joint service caching and computation offloading problem faces three significant challenges of dynamic tasks, heterogeneous resources, and coupled decisions. In this paper, we investigate the issue of joint service caching and computation offloading in mobile edge-cloud computing networks. Specifically, we formulate the optimization problem as minimizing the long-term average service latency, which is NP-hard. To solve the problem, we conduct in-depth theoretical analyses and decompose it into two sub-problems: service caching processing and computation offloading processing. We are the first to propose a novel hierarchical deep reinforcement learning algorithm to solve the formulated problem, where multiple edge agents and a cloud agent collaboratively determine the caching-action and offloading-action, respectively. The results obtained through trace-driven simulations reveal that the proposed framework outperforms several prevailing algorithms concerning the average service latency across diverse scenarios. In a complex real scenario, our framework achieves an approximately 33% convergence improvement and a remarkable 39% reduction in the average service latency when compared to reinforcement learning-based algorithms.

关键词： computation offloading hierarchical deep reinforcement learning mobile edge-cloud computing service caching computation offloading hierarchical deep reinforcement learning mobile edge-cloud computing service caching

来源：评论

学校读者我要写书评

暂无评论

Perception data fusion-based computation offloading in cooperative vehicle infrastructure systems

引用

JOURNAL OF SUPERCOMPUTING 2024年第12期80卷 17688-17710页

作者： Wu, Ruizhi Li, Bo Hou, Peng Hou, Fen Yunnan Univ Sch Informat Sci & Engn Kunming 650091 Yunnan Peoples R China Fudan Univ Sch Comp Sci & Engn Shanghai 200438 Peoples R China Univ Macau Dept Elect & Comp Engn Macau 999078 Peoples R China

With the rapid advancement of intelligent transportation systems, cooperative vehicle infrastructure systems emerge as a vital frontier for development. Real-time mutual fusion of perception data is a crucial technology for ensuring the security of ITS systems. However, the computation-intensive nature of perception data fusion poses a significant challenge in terms of computing resource allocation and scheduling. In this paper, we propose a vehicle-road cooperative network that facilitates computation offloading during the real-time perception data fusion process. We present an architecture that enables users to generate tasks and offload computations, and we formulate an integer nonlinear programming problem within this framework. Considering the dynamic, random, and time-varying characteristics of cooperative vehicle infrastructure systems, we introduced the Deep Deterministic Policy Gradient (DDPG) algorithm for perception fusion computing offloading (DDPG-PFCO). Through extensive experiments conducted on a real map, experimental results show that the proposed algorithm outperforms other comparison algorithms, exhibiting significant improvements in performance.

关键词： Intelligent transportation systems Perception data fusion computation offloading Cooperative vehicle infrastructure systems

来源：评论

学校读者我要写书评

暂无评论

Adaptive computation offloading for latency-sensitive tasks in heterogeneous edge-cloud-enabled smart warehouses using Gau-Angle FIS and AGE-MOEA-II

引用

WIRELESS NETWORKS 2024年第7期30卷 6493-6506页

作者： Zhao, Bohai Shen, Xinchun Peng, Kai Wang, Jiabin Leung, Victor C. M. Huaqiao Univ Coll Engn Quanzhou 362000 Peoples R China Nanjing Univ State Key Lab Novel Software Technol Nanjing 210000 Peoples R China Shenzhen Univ Coll Comp Sci & Software Engn Shenzhen 518060 Peoples R China Univ British Columbia Dept Elect & Comp Engn Vancouver BC V6T 1Z4 Canada

Industry 4.0 has introduced new development opportunities and ideological guidance to the structural reform of the supply side, as well as new requirements and challenges for the warehousing industry. Building a smart and effective warehouse system has become a crucial point for enterprises to gain a competitive advantage in market competition. As a conduit and link connecting all aspects of logistics, smart warehouses are confronted with increasing digital information and higher complexity and variability than traditional paradigms. In light of this, we establish an edge-cloud-enabled smart warehouse system and model the large-scale time-sensitive tasks based on directed acyclic graphs (DAGs). Then, by analyzing the service procedure and the cost of the devised system, an adaptive computation offloading method is proposed, which employs the Gau-Angle fuzzy inference system (FIS) for adaptive updating operators and achieves precise population screening based on the improved Pareto front modeling algorithm for large-scale many-objective optimization (AGE-MOEA-II), named ACOL-ECW. Finally, the effectiveness and superiority of ACOL-ECW are proved by comparative experiments conducted at various data scales and service scenarios.

关键词： Smart warehouse computation offloading Edge computing Time-sensitive tasks Gau-Angle FIS

来源：评论

学校读者我要写书评

暂无评论

Deep Neural Networks meet computation offloading in mobile edge networks: Applications, taxonomy, and open issues

引用

JOURNAL OF NETWORK AND COMPUTER APPLICATIONS 2024年 226卷

作者： Mustafa, Ehzaz Shuja, Junaid Rehman, Faisal Riaz, Ahsan Maray, Mohammed Bilal, Muhammad Khan, Muhammad Khurram COMSATS Univ Islamabad Dept Comp Sci Abbottabad 22060 Pakistan Univ Teknol PETRONAS Dept Comp & Informat Sci Seri Iskandar Malaysia King Khalid Univ Dept Informat Syst Abha Saudi Arabia Univ Lancaster Sch Comp & Commun Lancaster England King Saud Univ Riyadh Saudi Arabia

Mobile Edge Computing (MEC) is a modern paradigm that involves moving computing and storage resources closer to the network edge, reducing latency, and enabling innovative, delay-sensitive applications. Within MEC, computation offloading refers to the process of transferring computationally intensive tasks or processes from mobile devices to edge servers, optimizing the performance of mobile applications. Traditional numerical optimization methods for computation offloading often necessitate numerous iterations to attain optimal solutions. In this paper, we provide a tutorial on how Deep Neural Networks (DNNs) resolve the challenges of computation offloading. The article explores various applications of DNNs in computation offloading, encompassing channel estimation, caching, AR and VR applications, resource allocation, mode selection, unmanned aerial vehicles (UAVs), and vehicle management. We present a comprehensive taxonomy that categorizes these applications, and offer an overview of existing schemes, comparing their effectiveness. Additionally, we outline the open research issues that can be addressed through the application of DNNs in MEC offloading. We also highlight specific challenges related to DNN utilization in computation offloading. In conclusion, we affirm that DNNs are widely acknowledged as invaluable tools for optimizing computation offloading in MEC.

关键词： Mobile edge computing computation offloading Deep Neural Networks Supervised learning Unsupervised learning Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

An efficient computation offloading in edge environment using genetic algorithm with directed search techniques for IoT applications

引用

FUTURE GENERATION COMPUTER SYSTEMS-THE INTERNATIONAL JOURNAL OF ESCIENCE 2024年 158卷 378-390页

作者： Rajapackiyam, Ezhilarasie Devi, Anousouya Reddy, Mandi Sushmanth Arumugam, Umamakeswari Vairavasundaram, Subramaniyaswamy Vairavasundaram, Indragandhi Suresh, Vishnu SASTRA Deemed Univ Sch Comp Thanjavur India SRM Inst Sci & Technol Dept Computat Intelligence Kattankulathur India Amazon Dev Ctr Pvt Ltd Chennai India Vellore Inst Technol Sch Comp Sci & Engn Vellore India Vellore Inst Technol Sch Elect Engn Vellore India Wroclaw Univ Sci & Technol Fac Elect Engn PL-50370 Wroclaw Poland

The contemporary computations in the IoT environment are often outsourced to remote infrastructures due to inherent computation -intensive nature of tasks or to circumvent the high expenditures associated with establishing local infrastructures. Traditional approaches, be it cloud -based or localized, are deemed impractical for computation offloading in this scenario due to the trade-off between benefits and increased latency which is unacceptable for many real-time IoT applications. computation offloading at the edge environment is a promising solution in leveraging the untapped resources at Edge. offloading application execution to edge servers offers the potential to reduce completion time and fulfill application requirements. However, the simultaneous offloading of multiple entire applications to edge servers may strain their hardware and communication channels and also poses significant challenge for ensuring Quality of Service. To overcome this, a multi -site workflow offloading problem is conceptualized wherein tasks are strategically distributed across various edge sites for efficient execution. The problem is combinatorial and an optimal solution is never guaranteed through deterministic procedures. Thus this work proposes a novel metaheuristic approach based on modified Genetic Algorithm with Directed Search (GADS) operators that aim at optimizing the overall Execution Time (ET) and Energy Consumed (EC). The efficiency of GADS is demonstrated against alternatives, and a relative comparison is carried out in terms of the quality of the solution and the pace of convergence to the solution and approximately a minimum of 1% and maximum 15% gain is achieved.

关键词： computation offloading IoT Edge computing Genetic algorithm Task scheduling Multi-objective optimization

来源：评论

学校读者我要写书评

暂无评论

computation offloading in Mobile Edge Computing for Next Generation Networks: A deep reinforcement learning approach 2

Computation Offloading in Mobile Edge Computing for Next Gen...

引用

2nd International Conference for Advancement in Technology, ICONAT 2023

作者： Dhiman, Shrey Chauhan, Amita Kasuhal, Sakshi Kumar, Harish University Institute of Engineering and Technology Panjab University Chandigarh India

ISBN: (纸本)9781665475174

The next generation network 5G and beyond will provide higher speed, greater capability and lower latency for high-end technologies like augmented reality, online gaming, robotic arm surgery, high-quality video streaming, etc. Mobile Edge Computing (MEC) brings computing, storage and networking resources closer to the end user to host the compute-intensive and latency-sensitive applications at the edge of the network. Presently, the UAV-equipped Mobile Edge Computing (MEC) system provides computation services to mobile devices on the ground. However, the issue of processing delay and energy consumption in the task offloading process needs to be addressed. In the present paper, a novel Deep Deterministic Policy Gradient (DDPG) based approach is proposed that shall reduce the processing time by simultaneously improving user scheduling, resource allotment and UAV maneuverability, and formulating computation offloading problem as a high non-convex objective function. The performance of the suggested approach is demonstrated by simulation results using real-world parameters and the obtained results are compared to state-of-the-art algorithms. © 2023 IEEE.

关键词： computation offloading

来源：评论

学校读者我要写书评

暂无评论

Multi-agent reinforcement learning with graph representation for green edge-cloud computation offloading

引用

COMPUTER COMMUNICATIONS 2025年 239卷

作者： Bo, Yifan Feng, Jinghan Zhang, Shou Leng, Biao Beihang Univ Sch Comp Sci & Engn Beijing 100191 Peoples R China Beijing Jiaotong Univ Sch Comp & Informat Technol Beijing 100044 Peoples R China

Edge-Cloud Computing (ECC) stands as a widely adopted distributed computing architecture that facilitates offloading of computation-intensive tasks from Internet of Things (IoT) devices to edge servers. The growing emphasis on energy conservation and environmental protection raises the concerns of green edge-cloud computation offloading technology. However, conventional computation offloading methods have difficulties in making real-time offloading decisions and adapting to dynamic environmental changes, such as communication channels. In response to these challenges, we propose a multi-agent reinforcement learning method with graph representation to address the edge-cloud computing offloading schedule problem. Our approach constructs a multi-agent computation offloading reinforcement learning scenario and utilizes graph neural networks to represent the connectivity features between devices and edge servers. Experimental results demonstrate that our proposed method outperforms other algorithms in reducing system energy consumption and response delay. Furthermore, the time-consuming of our approach is significantly shorter compared heuristic genetic algorithms, with a reduction of 10-20 times.

关键词： Green edge-cloud computing computation offloading Multi-agent reinforcement learning Graph neural networks

来源：评论

学校读者我要写书评

暂无评论

computation offloading in Beyond 5G/6G Networks with Edge Computing: Implications and Challenges 20th

Computation Offloading in Beyond 5G/6G Networks with Edge ...

引用

20th International Symposium on Distributed Computing and Artificial Intelligence, DCAI 2023

作者： Stan, Catalina Rommel, Simon Vegas Olmos, Juan José Tafur Monroy, Idelfonso Department of Electrical Engineering Eindhoven University of Technology Eindhoven Netherlands NVIDIA Corporation Ofer Industrial Park Yokneam Israel

ISBN: (纸本)9783031383175

The emerging beyond 5G/6G networks come with novel, latency-sensitive and computation-intensive applications that require enhanced network performance and infrastructure to meet the expected quality of experience for end users. To cope with this challenge, computation offloading leverages the benefits of multi-access edge computing to migrate the application tasks requiring additional computing resources for reduced execution delay. Although the benefits of introducing offloading mechanisms into the network might be straightforward, the implementation is not trivial due to various communication and computation trade-offs that must be made to obtain optimal offloading decisions. In this paper, we provide an overview of computation offloading with highlight on the networking perspective by looking at different offloading decisions, current research efforts, as well as the challenges that may be encountered while building an efficient and robust offloading mechanism. In addition, we provide our view on the evolution of computation offloading in 6G networks to support novel applications through enriched infrastructure and powerful artificial intelligence techniques. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.

关键词： computation offloading

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：