A Peer-to-Peer multiway tree network with efficient range query search using multidimensional indexing

Multidimensional complex query searching is difficult in Peer-to-Peer (P2P) overlay networks due to its characteristics of high scalability, dynamicity and decentralized architecture. On the other side, the P2P applications like searching of multimedia contents in social networking or multiplayer games (that combines the search based on the multiple game scenarios) are becoming popular among the end users. However, due to the large amount of multidimensional data distributed among the thousands of peers, the support of complex queries like similarity or range search is essential as compared to the exact match queries. The existing literature (combining P2P networks and Multidimensional Indexing (MI)) either suffers from the poorer support to the range query or the range query search cost is limited to O(log₂N) or else it varies with the dimensionality. To address these problems, the traditional MI techniques based on the multiway tree structure (having larger fanout of the tree) can be employed to strengthen the multidimensional range query search capabilities. Based on our observations, a hybrid model combining the m-ary (m is fanout of the tree, m > 2) P2P tree network and MI based on the space containment relationships is preferred to reduce the range query search performance bound to O(log_mN). We show an illustrative model in this paper that takes advantages of P2P networks as well as MI based on mary tree to efficiently handle the range queries in the distributed multidimensional applications. However the main focus in our paper is to show how this model improves the search performance of the range queries (bound to O(log_mN) independent of the dimensionality of the multidimensional objects.