容错多智能体系统的分散体系结构

Z. Khan, S. Shahid, H. F. Ahmad, Arshad Ali, H. Suguri
{"title":"容错多智能体系统的分散体系结构","authors":"Z. Khan, S. Shahid, H. F. Ahmad, Arshad Ali, H. Suguri","doi":"10.1109/ISADS.2005.1452043","DOIUrl":null,"url":null,"abstract":"Multi agent systems (MAS) are expected to be involved in futuristic technologies. Agents require some execution environment in which they can publish their service interfaces and can provide services to other agent. Such execution environment is called agent platform (AP). From a technical point of view any abnormal behavior of platform can distress agents residing on that platform. That's why it is necessary to provide a suitable architecture for the AP which should not only provide fault tolerance but also scalability features. There also exist some management components within the platform, which provide services to application agents. All the agents within MAS are managed by agent management system (AMS) which is the mandatory supervisory authority of any AP. To be more scalable, a single agent platform can be distributed over several machines which not only provides load balancing but also fault tolerance depending upon the distributed architecture of the AP. In existing systems, AMS is centralized i.e. it exists on one machine. With centralized AMS, this infrastructure lacks fault tolerance, which is a key feature of high assurance. Absence of fault tolerance is the main reason for the small number of deployments of MAS. Failure of AMS leads towards abnormal behavior in the distributed platform. This paper proposes virtual agent cluster (VAC) paradigm which strongly supports decentralized AMS to achieve fault tolerance in distributed AP. VAC provides fault tolerance by using separate communication layers among different machines. Experiments show that it improves performance, brings autonomy and supports fault recovery along with load balancing in distributed AP.","PeriodicalId":120577,"journal":{"name":"Proceedings Autonomous Decentralized Systems, 2005. ISADS 2005.","volume":"214 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2005-04-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"31","resultStr":"{\"title\":\"Decentralized architecture for fault tolerant multi agent system\",\"authors\":\"Z. Khan, S. Shahid, H. F. Ahmad, Arshad Ali, H. Suguri\",\"doi\":\"10.1109/ISADS.2005.1452043\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Multi agent systems (MAS) are expected to be involved in futuristic technologies. Agents require some execution environment in which they can publish their service interfaces and can provide services to other agent. Such execution environment is called agent platform (AP). From a technical point of view any abnormal behavior of platform can distress agents residing on that platform. That's why it is necessary to provide a suitable architecture for the AP which should not only provide fault tolerance but also scalability features. There also exist some management components within the platform, which provide services to application agents. All the agents within MAS are managed by agent management system (AMS) which is the mandatory supervisory authority of any AP. To be more scalable, a single agent platform can be distributed over several machines which not only provides load balancing but also fault tolerance depending upon the distributed architecture of the AP. In existing systems, AMS is centralized i.e. it exists on one machine. With centralized AMS, this infrastructure lacks fault tolerance, which is a key feature of high assurance. Absence of fault tolerance is the main reason for the small number of deployments of MAS. Failure of AMS leads towards abnormal behavior in the distributed platform. This paper proposes virtual agent cluster (VAC) paradigm which strongly supports decentralized AMS to achieve fault tolerance in distributed AP. VAC provides fault tolerance by using separate communication layers among different machines. Experiments show that it improves performance, brings autonomy and supports fault recovery along with load balancing in distributed AP.\",\"PeriodicalId\":120577,\"journal\":{\"name\":\"Proceedings Autonomous Decentralized Systems, 2005. ISADS 2005.\",\"volume\":\"214 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2005-04-04\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"31\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Proceedings Autonomous Decentralized Systems, 2005. ISADS 2005.\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/ISADS.2005.1452043\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Proceedings Autonomous Decentralized Systems, 2005. ISADS 2005.","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ISADS.2005.1452043","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 31

摘要

多智能体系统(MAS)有望参与未来技术。代理需要一些执行环境,它们可以在其中发布自己的服务接口,并向其他代理提供服务。这种执行环境称为代理平台(AP)。从技术角度来看,平台的任何异常行为都会使驻留在该平台上的代理陷入困境。这就是为什么有必要为AP提供合适的体系结构,该体系结构不仅应该提供容错功能,还应该提供可伸缩性功能。平台中还存在一些管理组件,它们向应用程序代理提供服务。MAS中的所有代理都由代理管理系统(AMS)管理,该系统是任何AP的强制性监督机构。为了更具可扩展性,单个代理平台可以分布在多台机器上,这不仅提供负载平衡,还可以根据AP的分布式架构提供容错。在现有系统中,AMS是集中式的,即它存在于一台机器上。对于集中式AMS,这种基础设施缺乏容错性,而容错性是高保证的关键特征。缺乏容错能力是MAS部署数量少的主要原因。AMS故障会导致分布式平台出现异常行为。本文提出了虚拟代理集群(VAC)模式,该模式强有力地支持分散的AMS来实现分布式AP的容错。VAC通过在不同机器之间使用单独的通信层来实现容错。实验表明,该方法提高了分布式AP的性能,带来了自主性,并支持故障恢复和负载均衡。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
查看原文
分享 分享
微信好友 朋友圈 QQ好友 复制链接
本刊更多论文
Decentralized architecture for fault tolerant multi agent system
Multi agent systems (MAS) are expected to be involved in futuristic technologies. Agents require some execution environment in which they can publish their service interfaces and can provide services to other agent. Such execution environment is called agent platform (AP). From a technical point of view any abnormal behavior of platform can distress agents residing on that platform. That's why it is necessary to provide a suitable architecture for the AP which should not only provide fault tolerance but also scalability features. There also exist some management components within the platform, which provide services to application agents. All the agents within MAS are managed by agent management system (AMS) which is the mandatory supervisory authority of any AP. To be more scalable, a single agent platform can be distributed over several machines which not only provides load balancing but also fault tolerance depending upon the distributed architecture of the AP. In existing systems, AMS is centralized i.e. it exists on one machine. With centralized AMS, this infrastructure lacks fault tolerance, which is a key feature of high assurance. Absence of fault tolerance is the main reason for the small number of deployments of MAS. Failure of AMS leads towards abnormal behavior in the distributed platform. This paper proposes virtual agent cluster (VAC) paradigm which strongly supports decentralized AMS to achieve fault tolerance in distributed AP. VAC provides fault tolerance by using separate communication layers among different machines. Experiments show that it improves performance, brings autonomy and supports fault recovery along with load balancing in distributed AP.
求助全文
通过发布文献求助,成功后即可免费获取论文全文。 去求助
来源期刊
自引率
0.00%
发文量
0
期刊最新文献
An adaptive key-frame reference picture selection algorithm for video transmission via error prone networks Situational reasoning - a practical OWL use case The Yaoyorozu project designing ubiquitous information society in 2010 Autonomous decentralized resource allocation for tracking dynamic load change Maximizing the route reliability of mobile agents in unreliable environments
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
现在去查看 取消
×
提示
确定
0
微信
客服QQ
Book学术公众号 扫码关注我们
反馈
×
意见反馈
请填写您的意见或建议
请填写您的手机或邮箱
已复制链接
已复制链接
快去分享给好友吧!
我知道了
×
扫码分享
扫码分享
Book学术官方微信
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术
文献互助 智能选刊 最新文献 互助须知 联系我们:info@booksci.cn
Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。
Copyright © 2023 Book学术 All rights reserved.
ghs 京公网安备 11010802042870号 京ICP备2023020795号-1