Now a day millions of users to search and download desired data such as Napster and Gnutella as they are Peer-to-Pcer (P2P) files sharing applications. In order to improve performance in unstructured P2Ps replication strategies are used. Efficient and effective full-text retrieval over unstructured p2p networks was developed in order to address the problems of the query popularity independent replication strategies, previously a novel strategy, called Bloom Cast, which implements Bloom Filters in WP (With Pointers) scheme. In order to support random node sampling and network size estimation bloom Cast hybridizes a lightweight DHT with an unstructured P2Poverlay. However these well-organized techniques are executed irrespective of topologies and network size concerns. So, to overcome this problem we propose to use Compression bloom Filter, Long random walk, and Short Random Walk with local flooding schemes in accordance with variable p2p topologies and network sizes. Bloom filters these Hybrid Query Propagation schemes offers best performance over unstructured p2p networks and a practical execution authenticates the claim.
展开▼
