Research On Multi-Dimensional Data Storage And Query Based On P2P

OLAP(Online Analytical Processing) is a kind of decision supporting tool, which is based on data warehouse and supports multi-dimension analysis.OLAP is also an important means, and the user can acquire decision supporting from it. With the time passes by and the data's amount becomes larger and larger, the traditional technology of OLAP should be improved in order to meet the user's need of decision analysis. And now the technique of P2P continues to grow, we can use multi-dimensional data cube among some nodes of OLAP network to coordinate for finishing the decision analysis. Because not only it can improve efficiency of OLAP's query and analysis, but also can balance the OLAP server's load. So how to coordinate the nodes under the P2P environment becomes a key issue for the data warehouse in practice.Based on the aboving question, this paper establishes a kind of node's query under P2P environment, which provides its service by other OLAP's nodes. And the model that we absorb breaks the manner of providing service by a single server, which improve the efficiency of OLAP's decision analysis and balance the OLAP server's load. And this model is realized from the belowing three directions:(1) The storage model is based on the optimized multi-dimension data set and the trait of P2P environment. And it also makes use of the character of multi-dimension data set to store, which is according to the semantics dimension hierarchy chain and the extended semantics dimension hierarchy chain to compress the multi dimension data set. By this way, we can realize the share of multi-dimension data set among different nodes in the P2P environment.(2) Through establishing the framework of P2P network, the structure realizes the addition and exit of OLAP node dynamically. And this paper also proposes a kind of algorithm, which queries the data according to the multi-dimension data distributed query of the extend semantics chain. Because the algorithm realizes the data share of multiple nodes from the semantics layer in the P2P network, it whole improves query efficiency.(3) Based on the trait on P2P network environment and the analysis of multi-dimension data of the OLAP nodes, the model divides the nodes of OLAP into several parts according to whether their need is the same or not. Through establishing a virtual network, which is based on interest, the paper proposes an algorithm, which is based on the query analysis of OLAP's IADS. And this method can reduce the searching scope when OLAP is querying analysis. It also further optimizes the efficiency of multi-dimension data set's analysis in the P2P environment.