OID: Optimized Information Discovery Using Space Filling Curves in P2P Overlay Networks

Abstract

In this paper, we present the system design and evaluation of a Space Filling Curve (SFC)-based P2P information discovery system OID. The OID system uses multiple SFCs to significantly optimize the performance of multi-attribute range queries, particularly for applications with a large number of data attributes where a single big SFC-based index is inefficient. The basic idea is to have multiple SFCbased indices and select the best one to perform a query. We also introduce two tree-based query optimizations that increase the scalability of the system.

Index recommendation tool for optimized information discovery over distributed hash tables

“…The index with the least estimated cost is then used for actual resolution. Additional details of the OID system architecture can be found in our previous work [11].…”

“…The OID system uniformly distributes the identifier space of a SFC, [0, 2 k·d ), over the DHT [11]. Therefore, with a high probability, the number of peers responsible for query resolution increases with the number of queried zones and clusters.…”

“…A MAR query is resolved by estimating the performance of the query over each index, and then selecting the one with the best estimate. Although our approach outperforms the previously proposed approaches for extending DHTs [11] in terms of query overhead, the criterion for defining the initial set of indices is based on a simple heuristic. The OID system installs a user-defined number of indices for the most popular queries in the system.…”

“…Recently, we presented the Optimized Information Discovery (OID) system [11], which introduces a third type of indexing approach for extending DHTs to provide the support for MAR queries. The OID system creates a layer of multiple multi-attribute indices over a DHT.…”

“…This approach performs well if the queries include ranges over all data attributes (which is generally not the case for P2P information discovery systems). Data attributes that do not appear in queries are considered to be wildcards, and the performance of this approach deteriorates with increasing number of wildcards in queries [9], [11], [10].…”