Multiagent Systems Storage Resource Allocation in a Peer-to-Peer Distributed File System

Abstract

The objective of this research project is to understand and develop Multiagent Systems (MAS) storage resource allocation algorithms and methods where a peer-to-peer (P2P) system of computer resources is seen as a community of peers with different capabilities that collaborate to enhance the global performance in terms of storage resource balancing. The research focuses on P2P Distributed File Systems (DFS) storage resource allocation algorithms. Complex adaptive systems are evaluated and used to produce emergent global behaviours that can solve the storage resource allocation problem in a distributed system of peers. Squirrel behaviours provide a metaphor to develop algorithms and methods to allocate resources. A complete simulation software tool was developed to implement different behaviours that are analyzed on different scenarios and experiments. Experimental results support the initial hypothesis that hoarding mechaninsms found on squirrels behaviours allocate efficiently resources on a distributed system of consumers and providers of storage resources.