Task offloading of IOT device in fog-enabled architecture using deep reinforcement learning approach

IF 3 3区计算机科学 Q2 COMPUTER SCIENCE, INFORMATION SYSTEMS

Pervasive and Mobile Computing Pub Date : 2025-05-31 DOI:10.1016/j.pmcj.2025.102067

Abhinav Tomar, Megha Sharma, Ashwarya Agarwal, Aditya Nath Jha, Jai Jaiswal

{"title":"Task offloading of IOT device in fog-enabled architecture using deep reinforcement learning approach","authors":"Abhinav Tomar, Megha Sharma, Ashwarya Agarwal, Aditya Nath Jha, Jai Jaiswal","doi":"10.1016/j.pmcj.2025.102067","DOIUrl":null,"url":null,"abstract":"<div><div>The rapid growth of IoT devices has strained traditional cloud-centric architectures, revealing limitations in latency, bandwidth, and reliability. Fog computing addresses these issues by decentralizing resources closer to data sources, but task offloading and resource allocation remain challenging due to dynamic workloads, heterogeneous resources, and strict QoS requirements. This study models task offloading as a multi-objective optimization problem, considering task priority, energy efficiency, latency, and deadlines. Using a Markov Decision Process (MDP), it applies three Deep Reinforcement Learning (DRL) algorithms — DQN, DDPG, and SAC — in a multi-agent fog computing setup. Unlike prior work focused on single-agent or isolated metrics, this approach captures inter-node dependencies to improve overall resource use. Simulations show SAC achieves a 97.3% task deadline success rate and improves resource efficiency by 10.1%, highlighting its effectiveness in managing dynamic fog environments. These results advance scalable, adaptive offloading strategies for future IoT systems.</div></div>","PeriodicalId":49005,"journal":{"name":"Pervasive and Mobile Computing","volume":"112 ","pages":"Article 102067"},"PeriodicalIF":3.0000,"publicationDate":"2025-05-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Pervasive and Mobile Computing","FirstCategoryId":"94","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S1574119225000562","RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"COMPUTER SCIENCE, INFORMATION SYSTEMS","Score":null,"Total":0}

引用次数: 0

Abstract

The rapid growth of IoT devices has strained traditional cloud-centric architectures, revealing limitations in latency, bandwidth, and reliability. Fog computing addresses these issues by decentralizing resources closer to data sources, but task offloading and resource allocation remain challenging due to dynamic workloads, heterogeneous resources, and strict QoS requirements. This study models task offloading as a multi-objective optimization problem, considering task priority, energy efficiency, latency, and deadlines. Using a Markov Decision Process (MDP), it applies three Deep Reinforcement Learning (DRL) algorithms — DQN, DDPG, and SAC — in a multi-agent fog computing setup. Unlike prior work focused on single-agent or isolated metrics, this approach captures inter-node dependencies to improve overall resource use. Simulations show SAC achieves a 97.3% task deadline success rate and improves resource efficiency by 10.1%, highlighting its effectiveness in managing dynamic fog environments. These results advance scalable, adaptive offloading strategies for future IoT systems.

查看原文本刊更多论文

使用深度强化学习方法在雾支持架构中卸载物联网设备的任务

物联网设备的快速增长给传统的以云为中心的架构带来了压力，暴露出延迟、带宽和可靠性方面的局限性。雾计算通过分散离数据源更近的资源来解决这些问题，但是由于动态工作负载、异构资源和严格的QoS要求，任务卸载和资源分配仍然具有挑战性。本研究将任务卸载建模为一个多目标优化问题，考虑了任务优先级、能效、延迟和截止日期。使用马尔可夫决策过程（MDP），它在多代理雾计算设置中应用了三种深度强化学习（DRL）算法- DQN， DDPG和SAC。与之前关注单个代理或孤立度量的工作不同，该方法捕获节点间依赖关系，以提高整体资源使用。仿真结果表明，SAC算法的任务期限成功率为97.3%，资源效率提高了10.1%，在管理动态雾环境方面具有较好的效果。这些结果为未来的物联网系统提供了可扩展的、自适应的卸载策略。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Pervasive and Mobile Computing COMPUTER SCIENCE, INFORMATION SYSTEMS-TELECOMMUNICATIONS

CiteScore

7.70

自引率

2.30%

发文量

审稿时长

68 days

期刊介绍： As envisioned by Mark Weiser as early as 1991, pervasive computing systems and services have truly become integral parts of our daily lives. Tremendous developments in a multitude of technologies ranging from personalized and embedded smart devices (e.g., smartphones, sensors, wearables, IoTs, etc.) to ubiquitous connectivity, via a variety of wireless mobile communications and cognitive networking infrastructures, to advanced computing techniques (including edge, fog and cloud) and user-friendly middleware services and platforms have significantly contributed to the unprecedented advances in pervasive and mobile computing. Cutting-edge applications and paradigms have evolved, such as cyber-physical systems and smart environments (e.g., smart city, smart energy, smart transportation, smart healthcare, etc.) that also involve human in the loop through social interactions and participatory and/or mobile crowd sensing, for example. The goal of pervasive computing systems is to improve human experience and quality of life, without explicit awareness of the underlying communications and computing technologies. The Pervasive and Mobile Computing Journal (PMC) is a high-impact, peer-reviewed technical journal that publishes high-quality scientific articles spanning theory and practice, and covering all aspects of pervasive and mobile computing and systems.