QueTy:支持时间边际效益的云服务协同框架

发布时间：2018-03-30 19:36

  本文选题：云服务　切入点：多agent系统　出处：《计算机工程与应用》2017年19期

【摘要】：在不断衍化的互联网环境下,云服务呈现出异构分散,动态多变,数据规模庞大等特征,结合具有自主决策能力的多agent系统(MAS)来实现云服务的分配和管理近年来逐渐成为研究热点。为追踪整体效用随时间的衰减趋势从而进一步优化协同策略,在云服务Saa S层基础上,提出基于时间边际效益的协同框架——Que Ty。该框架为云服务的broker(代理)、provider(提供商)和executor(服务执行者)之间引入时间边际效益概念,并在此基础上提出通信衰减算法和选播算法,实现了对服务流程的控制,并进一步平衡了整体效用,降低了通信损耗。
[Abstract]:In the constantly evolving Internet environment, cloud services show the characteristics of heterogeneous, dynamic, dynamic, large data scale, and so on. In recent years, it has become a hot topic to realize cloud service allocation and management by combining multi-#en0# system with autonomous decision ability. In order to track the decreasing trend of overall utility over time, the cooperative strategy can be further optimized. On the basis of cloud service Saa S layer, this paper proposes a collaborative framework based on temporal marginal benefit, named que Ty.The framework introduces the concept of time marginal benefit between broker (broker) and executor (service executor) of cloud service. On this basis, the communication attenuation algorithm and the anycast algorithm are proposed to control the service flow, and the overall utility is further balanced, and the communication loss is reduced.
【作者单位】： 北京信息科技大学计算机学院;
【基金】：“十二五”国家科技支撑计划课题(No.2014BAH25F03) 国家自然科学基金(No.61402043)
【分类号】：TP393.09
，

本文编号：1687341