Meta-reinforcement learning driven model architecture and algorithm optimization in intelligent driving task offloading

IF 4.3 3区计算机科学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Computer Communications Pub Date : 2025-08-19 DOI:10.1016/j.comcom.2025.108310

Peiying Zhang , Jiamin Liu , Zhiyuan Ren , Lizhuang Tan , Neeraj Kumar , Konstantin Igorevich Kostromitin

{"title":"Meta-reinforcement learning driven model architecture and algorithm optimization in intelligent driving task offloading","authors":"Peiying Zhang , Jiamin Liu , Zhiyuan Ren , Lizhuang Tan , Neeraj Kumar , Konstantin Igorevich Kostromitin","doi":"10.1016/j.comcom.2025.108310","DOIUrl":null,"url":null,"abstract":"<div><div>In the process of rapid development of intelligent driving technology, the amount of data generated by vehicles increases dramatically, while the bottleneck of storage and computation capacity of in-vehicle devices becomes more and more prominent, and task offloading becomes the key to improve the performance of intelligent driving systems. In this context, this paper proposes the MRL-ADTO algorithm, which innovatively applies meta-reinforcement learning (MRL) to the field of intelligent driving task offloading, optimizes the directed acyclic graph (DAG) synthesis logic and the task priority ranking algorithm, designs a neural network model based on the sequence to sequence (Seq2Seq) structure, and introduces the mechanism of multi-head attention at the same time. The experimental results show that MRL-ADTO can significantly reduce the task execution delay in multiple scenarios compared with the existing algorithms, and has obvious advantages in terms of training efficiency and convergence performance, providing an efficient and reliable solution for smart driving task offloading.</div></div>","PeriodicalId":55224,"journal":{"name":"Computer Communications","volume":"242 ","pages":"Article 108310"},"PeriodicalIF":4.3000,"publicationDate":"2025-08-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Computer Communications","FirstCategoryId":"94","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S0140366425002671","RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"COMPUTER SCIENCE, INFORMATION SYSTEMS","Score":null,"Total":0}

引用次数: 0

Abstract

In the process of rapid development of intelligent driving technology, the amount of data generated by vehicles increases dramatically, while the bottleneck of storage and computation capacity of in-vehicle devices becomes more and more prominent, and task offloading becomes the key to improve the performance of intelligent driving systems. In this context, this paper proposes the MRL-ADTO algorithm, which innovatively applies meta-reinforcement learning (MRL) to the field of intelligent driving task offloading, optimizes the directed acyclic graph (DAG) synthesis logic and the task priority ranking algorithm, designs a neural network model based on the sequence to sequence (Seq2Seq) structure, and introduces the mechanism of multi-head attention at the same time. The experimental results show that MRL-ADTO can significantly reduce the task execution delay in multiple scenarios compared with the existing algorithms, and has obvious advantages in terms of training efficiency and convergence performance, providing an efficient and reliable solution for smart driving task offloading.

查看原文本刊更多论文

元强化学习驱动的智能驾驶任务卸载模型架构及算法优化

在智能驾驶技术快速发展的过程中，车辆产生的数据量急剧增加，而车载设备存储和计算能力的瓶颈日益突出，任务卸载成为提高智能驾驶系统性能的关键。在此背景下，本文提出了MRL- adto算法，该算法创新性地将元强化学习（MRL）应用于智能驾驶任务卸载领域，优化了有向无环图（DAG）合成逻辑和任务优先级排序算法，设计了基于序列到序列（Seq2Seq）结构的神经网络模型，同时引入了多人关注机制。实验结果表明，与现有算法相比，MRL-ADTO可以显著降低多场景下的任务执行延迟，在训练效率和收敛性能方面具有明显优势，为智能驾驶任务卸载提供了高效可靠的解决方案。

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

Computer Communications 工程技术-电信学

CiteScore

14.10

自引率

5.00%

发文量

397

审稿时长

66 days

期刊介绍： Computer and Communications networks are key infrastructures of the information society with high socio-economic value as they contribute to the correct operations of many critical services (from healthcare to finance and transportation). Internet is the core of today''s computer-communication infrastructures. This has transformed the Internet, from a robust network for data transfer between computers, to a global, content-rich, communication and information system where contents are increasingly generated by the users, and distributed according to human social relations. Next-generation network technologies, architectures and protocols are therefore required to overcome the limitations of the legacy Internet and add new capabilities and services. The future Internet should be ubiquitous, secure, resilient, and closer to human communication paradigms. Computer Communications is a peer-reviewed international journal that publishes high-quality scientific articles (both theory and practice) and survey papers covering all aspects of future computer communication networks (on all layers, except the physical layer), with a special attention to the evolution of the Internet architecture, protocols, services, and applications.