Research Of Resource Discovery Mechanism In Mobile Grid

Grid connects broad geographical distributed resources (including computing resources, storage resources, communication resources, software resources, data resources, information resources, knowledge resources, external equipment, etc.) together into a logical infrastructure, to implement resource sharing and collaboration in the virtual network environments, eliminate isolated islands of information and resources, so it has great research value.The thesis focuses on the "Mobile Grid" which is composed by mobile devices, and dedicates to solve the Mobile Grid resource discovery mechanism. The Mobile Grid has highly dynamic, weak connectivity, restrictions on mobile devices and limited bandwidth characteristics, which make the existing grid that based on the centralized and hierarchical resource discovery mechanism is no longer applicable to the Mobile Grid. Good solution for Mobile Grid resource discovery mechanism, this thesis draws on P2P technology for integrated P2P resource discovery mechanism in Mobile Grid. The main research work and innovative in the following areas:Firstly, specialty for Mobile Grid system, this thesis presents a method to use the distributed and unstructured P2P architecture to organize the Mobile Grid resources. In this framework, there is no concept of client and server, each node is equal, is a provider of resources, but also a resource user. The program in this thesis is self-organized and scalable.Secondly, proposed an integrated P2P resource discovery model for Mobile Grid. In this model, the Mobile Grid nodes self-organized into two-tier P2P overlay networks:"Common network layer" and "Analysis of the network layer". Each of Mobile Grid node should be added into the "Common network layer", while some of the more stable performance and storage space nodes have free choice to added "Analysis of the network layer", to provide state analysis.Thirdly, on the basis of flooding algorithm, we proposed the Improved Local Indices algorithm. The Improved Local Indices, considering that the mobile terminal equipment has limited storage and limited network bandwidth, and for the highly dynamic of Mobile Grid environment, topology changes dramatically leading to node index information is failure, we use the macro mobility management and micro mobility management as the way to manage the movement of Mobile Grid environment. When the dynamics environment is low, resources position change is not dramatic, use of macro mobility management, to ensure the accuracy position information; when the environment is very frequent dynamic changes, the resource position dramatically, use of micro-mobility management, for each node to provide the resources position information as accurately as possible.Finally, established simulation environment under the NS2, to simulated Improved Local Indices algorithm, and compared with the classical flooding algorithm, further to test the effectiveness of this resource discovery algorithm.This thesis is supported by National Natural Science Foundation of China (No: 60970064,60773211), Open Fund of the State Key Laboratory of Software Development Environment (No:SKLSDE-2009KF-2-02), New Century Excellent Talents in university (No:NCET-08-0806), Fok Ying-Tong Education Foundation for Young Teachers in Higher Education Institutions of China (No:121067), the Topping Youth Foundation of Hubei Province (No:2008CDB335) and NSF of Wuhan Municipality (No:201010621207).