Despite their growing popularity, the behavior of super-peernetworks is not well understood. For example, what are the potential drawbacks of super-peer networks? How can super-peers be made more reliable? How many clients should a superpeer take on to maximize efficiency? In this paper we examinesuper-peer networks in detail, gaining an understanding of their fundamental characteristics and performance tradeoffs. We alsopresent practical guidelines and a general procedure for the design of an efficient super-peer network.

Abstract — In this paper, we propose a variant of Chord system that is resilient to routing attacks. By routing attacks, we mean the attacks which detour the looking up messages, aiming to disrupt the performance of data look-up systems by increasing the path length of queries. Chord routes messages uni-directionally and has no bi-directional edges. While its performance in the absence of routing attacks is acceptable, it degrades dramatically under routing attacks. Following this observation, we introduce the concept of reverse edges to the Chord system. We named the new Chord system as RChord. We propose several deterministic and randomized algorithms to construct reverse edges. We design a routing algorithm for the new system, which is simple, efficient and backward compatible with the original system. We then analyze the performance of the RChord under routing attacks. We find that its performance is significantly improved in terms of average path length, even by adding very few reverse edges. I.