Performance evaluation of advanced routing algorithms for unstructured peer-to-peer networks

Peer-to-peer systems have recently emerged to address the problem of enabling the virtualization of distributed resources such as processing, network bandwidth and storage capacity, to create a single system image, granting users and applications seamless access to vast IT capabilities. Participants in peer-to-peer networks are not only potential consumers but also potential resource providers, and operate autonomously with no central authority. Efficient resource sharing and discovery mechanisms are both essential for the functioning of the system as a whole and for the benefit of all participants.This paper illustrates our contributions to the characterization of unstructured peer-to-peer architectures, in which the overlay network topology and the adopted routing strategy are not deterministically correlated. Starting from classic analytical results in the field of random graphs, we introduce several novel topological models which put the emphasis on capturing the network growth, and that in our view are very significant for peer-to-peer systems. Moreover, we introduce a novel routing algorithm called HALO, which has been compared to the SRDI strategy adopted by JXTA. Simulation results of HALO and JXTA performance are provided for different overlay network topologies.