Multi-dimensional Queries in DHT-Based Peer-to-Peer Systems

Abstract

Supporting relational query processing in P2P data management systems needs multi-dimensional exact match queries processing and multi-dimensional range queries processing. The paper proposes a method for using a DHT-based P2P system to support multi-dimensional data while allowing range queries. The proposed system used a two-layered architecture, one for resource sharing and another for resource finding, so as to facilitate multi-dimensional query processing. The approach depends on using an order-preserving hash function and a relatively small number of P2P nodes to store data. Data is assigned to nodes by reducing its dimensionality to just one dimension, and is then mapped onto the P2P nodes, which are spread along the range of values of the one-dimensional data. After that, a multi-dimensional query algorithm (including to exact match query and range query) based on this strategy are proposed. At last, the paper offered experimental evidence to support the approach.