HRS: A Hybrid Replication Strategy for Exhaustive P2P Search

Successful search and versatile query support are two important requirements for peer-to-peer (P2P) search applications. Replication strategy is an effective approach to improve the search performance of unstructured P2P systems. However, existing replication strategies either adapt only to popular queries or incur excessive replication cost for unpopular queries. In this work, we propose HRS, a hybrid replication strategy to improve the search performance of unstructured P2P networks. By combining a query popularity independent strategy with the square-root strategy, HRS can effectively and efficiently handle both kind of queries, popular or not. We evaluate this design through mathematical proof and comprehensive simulations. Results show that HRS outperforms existing replication-based search paradigms in terms of search performance and resource consummation.