Range Queries in Massively Distributed Data

Abstract

Efficient support of semantic location queries in distributed hash table (DHT) P2P systems is a key challenge. It is difficult to handle since object distribution criteria among data providers are not related to semantic criteria. This paper focus on range queries which are particularly expensive or impose strong restrictions in current proposals. We propose meta-data indexing to support queries involving inequalities and "like" operators in a scalable way. The proposed solution is portable, it can be used on top of any current DHT P2P system