Fuzzy searching and routing in unstructured mobile peer-to-peer networks

Abstract

Peer-to-Peer (P2P) networks offer a scalable solution for efficient query searching and sharing across the unstructured networks. With an increased overhead due to large amount of object searching and routing in unstructured P2P networks, it is a challenge to continue QoS routing among different mobile peers. Most existing mobile P2P protocols focus on inflexible techniques to route queries and discover objects of interest. Such common techniques incur a relatively high search time due to remarkable network traffic and duplication of query messages. The correlation between routing and mobility is crucial to efficiently search and route the query object in an overlay to avoid unnecessary consumption of network resources. Our previous fuzzy search controller model (Shah and Kim in 12th IEEE international conference on dependable, autonomic and secure computing (DASC), 2014) reduced the search time query processing in P2P networks, but it caused low hit rate and a high overhead due to peer mobility. Thus, this article proposes a scalable fuzzy controller based on probabilistic walk for unstructured mobile P2P networks to reduce the search time with controlled mobility. The search time is reduced by jumping a query walker to a 2-hop away ultrapeer, selected through a fuzzy scheme. Furthermore, each mobile ultrapeer shares its pong cache with its directly connected ultrapeer in order to increase the hit rate and reduce the network overhead. Simulations show that the fuzzy search controller gives better performance than the competing protocols in terms time 10% reduction in response time and 15% increase in hit rate in different mobility scenarios.