Emulation Research On The Technology Of Resource Scheduling In Grid Computing Based On Ant Colony Algorithm

Grid computing is a kind of perfect computing technology will be used in field of science computing,engrineering computing and some other large-scale computings. Grid core service is an important part of grid computing. As a centrum of entire grid computing,grid core service connect grid understratum groupwares to top groupwares and take charge entire grid system in order to ensure grid system works effectively.It is important to research grid core service technology. Resource scheduling technology is a part of grid core service technology. An advanced resource scheduling policy in favor of reducing total run time and total expense in grid computing, by this means enhance entire grid system's capability. Resource scheduling policy is a key technology and very important to grid computing.Grid computing is a kind of distributed computing. But the environment of grid computing is more complex than the environment of distributed computing, so the resource scheduling policy which design on distributed computing unsuitable the environment of grid computing. For example, traditional policy unable forecast resources' dynamic charge, think over the confines of resources,jobs, the cost of finish the jobs and some other problems in grid computing.ACA(Ant Colony Algorithm)^[45] is a sort of heuristic search algorithm. ACA simulate behavior that the ants search route from nest to food. This algorithm has many strongpoints, such as characteristics of positive feedback, distributing and robustness. ACA is fit for deal with combination-optimize questions. Resource scheduling is a combination-optimize question, on this account, thesis design a new resource scheduling policy which is suit for the environment of grid computing based on principle of ACA.We think carefully about the characteristics of grid computing in new policy. The aim of this new policy is schedule resources more in reason and improve capability of grid computing by this means.Thesis mostly expatiate the preparation,design process and testing about new resource scheduling policy which based on principle of ACA. The methods are discussed as below:(1) Parse the environment of grid computing. Definitude the characteristics of grid computing and sum up the problems which resource scheduling policy must resolve intoin grid computing. Summarize the reasons that traditional resource scheduling policy unsuit for the grid computing. Dwell on the principle and characterixtics of ACA. Bring forward the scheme that resolve problem about resource scheduling based on ACA and analyse the feasibility of this scheme.(2) Parse the standard ACA. Building testing programme according principle of standard ACA in order to find some factors that affect ACA capability through this programme. The testing result will provide a useful reference to design resource scheduling policy based on ACA.(3) Building grid simulation system based on GridSim toolkit and ameliorate standard system in order to suit for the run of ACA. Scheduling policy must reference any information about simulation system when policy make decision,so we must test some parameters of simulation system after build it. We test simulation system in order to find factors that affect system capability and how they affect each other.(4) Design resource scheduling policy based on ACA. We will reference the essence characteristics of ACA and the infection of the grid computing. The aim of the new policy is reduce the total run time and total cost through schedule resources more in reason, hi this way,we can improve capability of the entire grid system.(5) Test the performance of new resource scheduling policy through GridSim grid emulator. We find there aer two problem in new policy through testing. One is the new policy can't use resources sufficiently,the other is astringency of ACA is not very well.We analyse these problems and ameliorate resource scheduling policy.In the end of this work, we can prove the effectivity of above the new resource scheduling policy that we designed throuth GridSim emulator.