Research Of P2P Multi-dimension Data Searching Technique

With the development of computer network and communication technology, Peer-to-Peer (P2P) has become an important architecture of the network computing environment and distributed processing system on high performance computing area. Although that P2P network existing now have their own characteristic, but there is still none kinds of P2Ps' method of coming together could be meet the comprehensive demand that should be have in many ways such as function, performance, servicing quality etc. Although unstructured P2P system could support query multi-dimensioned if adopt the blind researching mechanism such as resemble flooding. Both the efficient of researching and the expansibility are lower. Especially both the researching veracity and the expansibility are important questions that unstructured confronted with should be solved.According to the purpose above, this thesis is mainly combine traditional multi-dimensional index technique and P2P system together, put forward a kind of novel multi-dimensional data index structure, give out the arithmetic for multi-dimensional data range query and the establishment and maintenance of index. To data search, this thesis has studied the technique of multi-dimensional data space indexes, put forward a comprehensive frame apply to the P2P network which have super node aimed to unstructured P2P system. In this frame, many processing can be carried out for the multi-dimensional data operations such as share, index and search etc. Bathed on R*-tree index tree, this thesis has putted forward a kind of expanded R*-tree index tree can be applied in P2P, named ER*-tree, it is a kind of expanded index R*-Tree which has the biggest space range. This frame can respond to the query of multi-dimensional data completely or similarly. In multi-dimensional data gathers, with the ER*-Tree as elements, apply range query processing arithmetic into the P2P system, to support complicated data search of multi-dimensional.This method can meet the needs basically to the resources detection of P2Ps and the commonness search of multi-dimensional, But for the management of data that according to the semantic query(text originally inspectional) of P2Ps is the next goal that we working for.