A near-optimal one hop DHT lookup in structure peer-to-peer networks

The distributed hash tables (DHTs) exhibit several interesting properties, such as self-configuration, self-maintenance, scalability and robustness that are clearly desirable for many kinds of applications in the Internet. However, Most of DHT approaches have high resolution latency because each lookup need contact several servers in sequence, typically O(log n) hops (n is the number of peers in the system), which is impossible practical for the Internet. In this paper, we present a near-optimal one hop lookup in structure peer-to-peer networks, which is able to achieve the maximal performance of the system with reasonable maintenance traffic overhead even for huge and dynamic networks. We design a fast and efficient event (such as peers join or leave the system) notification algorithm to maintain the complete network topology with the acceptable bandwidth usage. We evaluate the performance of the proposed approach in terms of effectiveness and feasibility. The analysis and simulation results show that even for very large systems the scheme has good scalability and low resolution delay.