State-based search strategy in unstructured P2P

As a major pattern for design of future heterogeneous Cloud and service-oriented networks, research on P2P networking is extremely important and could possibly significantly alter the way of implementation of large-scale of Internet applications - Cloud Computing. However, we will not reap all the benefits of utilising the distributed resources in future P2P Clouds, unless we have an efficient and scalable way to discover the desirable services regarding users' needs. Through analysis of peer node's multi-dimensional quality-of-service (QoS) attributes for resource discovery, it is found that peer node state's information is essential for improving search performance in distributed systems. This paper presents a novel efficient search algorithm, named State-Based Search (SBS), according to peer node's state information. This algorithm has been simulated and evaluated in dynamic P2P environments. By comparing other related interest-based algorithms, the experimental results show that the SBS algorithm is able to effectively improve performance by reducing search response time and achieving better load balance. (C) 2011 Elsevier B.V. All rights reserved.