Autonomic peer-to-peer systems: Incentive and security issues

Abstract

With voluntary users participating in an autonomic manner, peer-to-peer (P2P) systems have been proliferating in an unprecedented pace. Indeed, it is widely known that P2P traffic now constitutes over 60% of total Internet traffic. P2P systems are now used for file sharing, media streaming, and various other social networking applications. Furthermore, P2P systems are also extending their reach to the wireless realm. However, there are still two major system aspects that pose challenges to P2P systems' designers and users: incentives and security. First and foremost, a P2P system, by its nature, is viable only if users contribute their resources to the community. Obviously, uniform and global altruistic behaviors cannot be expected for all users. Some users will definitely try to take advantage of the altruism of other participants. If such free-riding phenomenon is too wide spread, then a system collapse will result because the input to the P2P community is smaller than the output. Thus, it is important to incorporate effective incentive mechanisms to deter selfish behaviors and encourage active contributions. On the other hand, security related issues such as privacy, anonymity, and authentication are also beyond doubt critical concerns of the participating users. Essentially, users do not want to sacrifice security to trade for service. Clearly security issues and incentives are closely related in that a low security or untrustworthy P2P community will not attract a large population of contributing users, and instead, might even tempt malicious users to consider the system as a potential point of attacks. In this chapter, we survey and analyze the current state-of-the-art in tackling the incentive and security issues in P2P systems. We first give a brief account of P2P applications and their wired and wireless operating environments. We then survey and critique existing incentive techniques. This is followed by the analysis of contemporary P2P security algorithms. The techniques that we survey will range from traditional optimization algorithms to game theoretic schemes. We provide some of our thoughts on open research issues, followed by a conclusion. © 2009 Springer Science+Business Media, LLC.