MP2P Resource Location Method On Node Capability And Resource Index

With the rapid development and wide application of the P2P network technology, as wellas the maturity of mobile network environment, the P2P network technology in the Internethas extended to the wireless and mobile devices, and will be more widely used and developedin the wireless and mobile environment. The MP2P network has become a hot research andapplication topic in the new generation network applications for strong distribution and wideparticipation. The research and industrialization on MP2P has important application value andpractical significance in many aspects of the field such as the military communications andthe new generation network communications. The resource location method is the foundationand core of the MP2P network researches and applications, which is not only contains thepropagation operation of the query request between nodes (i.e. the resource search), but alsoincludes the query and retrieval of the resource or its index information on a single node inthe location process. However, current methods exist to varying degrees with inadequateconsiderations for the node’s capacity and mobility, lower position efficiency, limited rangeof applications, etc. How to locate the necessary resource for the user quickly and accuratelyby using appropriate method within the limited time and the bandwidth range will greatlyaffect the efficiency and performance of the MP2P applications. Therefore, the resourcelocation method is one of the important challenges required to be solved in the MP2Pnetwork researches and applications.This thesis proposes the MP2P resource location model and method on the nodecapability and the resource index, which tightly rounds the characteristics of the MP2Pnetwork environment, makes the use of network resource as the core and directly faces thedemanded resources. Firstly, this thesis defines the value of the node capability and theresource index credibility as metrics, and studies their influence factors and change law,which provide the support of the foundation parameters for building a dynamic networkconnectivity model facing to the shared resource. Secondly, a resource index structure isdefined to reflect the connection relationship of the nodes. The structure is based on theresource information, and provides structural support for the model. Thirdly, a dynamic network connectivity model adapting to the MP2P network environment is built, which isbased on the value of the node capability and makes the resource index as the core. Themodel lays the foundation for achieving the efficient resource location. Fourthly, based on themodel above and the distribution characteristics of the demanded resource index, the networkis divided into the invalid area and the effective area of the resource index. In the invalid areaof the resource index, based on the values of the neighbor nodes capability and its flowcontrol flag, some different paths are selected by using biased random k-way search methodto forward the query message. Meanwhile a caching mechanism of the responded message isintroduced to assist the resource location, which accelerate the speed that query messagegetting into the effective area of the resource index. After the query message goes into theeffective area of the resource index, it will be forwarded along the reverse diffusion path ofthe resource index, and then the required resource will be hit quickly and accurately. Thus theefficient resource location has been implemented. Finally, the performance of the resourcelocation method mentioned above has been analyzed and evaluated with more detailsrespectively from the success rate aspect, the average positioning delay aspect and thenetwork load balancing aspect of the resource location. The method has a higher success rate,a lower positioning delay and a better network load balancing performance.