A recent survey revealed that 18 millions American Internet users, or approximately 14% of total American Internet population have peer-to-peer (P2P) file-sharing applications running on their computers (Rainie & Madden, 2004). Not surprisingly, P2P applications have become common tools for information sharing and distribution since the appearance of Napster (Napster, 2003) in 1999. P2P systems are the distributed systems in which all nodes are equal in terms of functionality and able to directly communicate with each other without the coordination of a powerful server. Anonymity, scalability, fault resilience, decentralization and self-organization are the distinct characteristics of P2P computing (Milojicic et al., 2002) compared with the traditional client-server computing. P2P computing is believed to be capable of overcoming limitations of the computing environment placed by the client-server computing model. Milojicic et al. (2002), for example, suggested that P2P computing is capable of providing improved scalability by eliminating the limiting factor, the centralized server existing in the client-server computing. In the past few years, P2P computing and its promised characteristics have caught the attention of researchers who have studied the existing P2P networks, and the advantages and disadvantage of P2P systems. Important findings include the excessive network traffic caused by flooding-based searching mechanism that must be tackled in order to fully utilize the improved scalability of P2P systems (Matei, Iamnitchi, & Foster, 2002; Portmann & Seneviratne, 2002). There were proposed efficient searching techniques targeted for both structured and unstructured P2P systems. Other research projects were conducted to study, and were intended to complement, the drawbacks brought by distinct characteristics of P2P systems. For example, the P2P users’ free-riding behavior is generally attributed to the anonymity of such form of communication (Adar & Huberman, 2000). Recent research projects have shifted to a new line of investigation of P2P networks from the economic perspective and applications of P2P systems in workplaces (Kwok & Gao, 2004; Tiwana, 2003).
P2P computing for intelligence of things
