基于联合意图的网格资源分配模型

Grid resource allocation model based on joint intentions

分析了网格资源管理过程中经典的资源分配模型,针对网格资源分配过程中的特点,构建了资源需求型Agent、资源提供型Agent、资源协调型Agent以及交互型Agent,建立了基于联合意图的网格资源分配模型.并给出了网格资源分配过程中的协商协议和协商算法,该算法利用多个Agent之间的共有目标进行交互,增强了问题求解的能力.此外,在传统网格资源管理体系结构的基础上,建立了基于多Agent的网格资源管理体系结构.在仿真平台下进行了对比实验,实验结果表明,利用新模型能够通过4种Agent角色之间的相互协商,合理地分配任务,提高网格资源的利用率.

The classic model of the resource allocation problem during the grid resource allocation process is analyzed.According to the characteristics of the process,four different Agent roles are proposed,including resource required Agent,resource supplied Agent,coordination Agent,and interactive Agent.Then a new grid resource allocation model is built based on joint intentions,and the coordinated negotiation protocol and the negotiation algorithms are also suggested.These algorithms can improve the capability for solving the problem.Furthermore,a multi-Agent based architecture for grid resource allocation is proposed on the basis of classical architecture.The experiments are done on the simulation platform.The results show that tasks are allocated effectively to the grid resources via negotiation between different Agent roles under the new model.