Neighbor Selection in Peer-to-Peer Overlay Networks: A Swarm Intelligence Approach

Abstract

Summary. Peer-to-peer (P2P) topology has a significant influence on the performance, search efficiency and functionality, and scalability of the application. In this chapter, we investigate a multi-swarm approach to the problem of Neighbor Selection (NS) in P2P networks. Particle swarm share some common characteristics with P2P in the dynamic socially environment. Each particle encodes the upper half of the peer-connection matrix through the undirected graph, which reduces the search space dimension. The portion of the adjustment to the velocity influenced by the individual's cognition, the group cognition from multi-swarms, and the social cognition from the whole swarm, makes an important influence on the particles' ergodic and synergetic performance. We also attempt to theoretically prove that the multiswarm optimization algorithm converges with a probability of 1 towards the global optima. The performance of our approach is evaluated and compared with other two different algorithms. The results indicate that it usually required shorter time to obtain better results than the other considered methods, specially for large scale problems.

Lonet

“…In addition, researchers [Ghanea-Hercock et al 2006;Koo et al 2006] posed the issue on neighborhood selection by assigning different weights to nodes. The Hebbian rule was applied in the weighted network to revise the swarm intelligence algorithm for optimal neighborhood selection within a social-like peer-to-peer network [Liu et al 2009]. …”