Biology-Inspired Optimizations of Peer-to-Peer Overlay Networks

Abstract

The aim of this article is to examine the relationship of large-scale Peer-to-Peer (P2P) overlay networks and certain biological systems. In particular, we focus on organization mechanisms that are crucial to adjust and optimize the behavior of large-scale P2P systems in the face of a dynamic environment. We propose to adopt concepts and mechanisms of biological systems in order to extend their capabilities to cope with environmental changes, e.g. a highly dynamic network topology. We introduce the notion of organic P2P overlay networks that adopt behavioral and structural characteristics of biological systems. We present a framework that poses as a basis for understanding, investigating, and implementing organic P2P overlay networks. Using a case study, we analyze an organic P2P overlay network, AntCAN, that utilizes ant colony optimization to improve the query processing in the face of varying query distribution patterns. Experiments confirm the functional efficiency of this self-organization mechanisms as well as the applicability of our proposed framework.