Uinta: A P2P Routing Algorithm Based on the User's Interest and the Network Topology

Peer-to-peer (P2P) overlay networks, such as CAN, Chord, Pastry and Tapestry, lead to high latency and low efficiency because they are independent of underlying physical networks. A well-routed lookup path in an overlay network with a small number of logical hops can result in a long delay and excessive traffic due to undesirably long distances in some physical links. In these DHT-based P2P systems, each data item is associated with a key and the key/value pair is stored in the node to which the key maps, not considering the data semantic. In this paper, we propose an effective P2P routing algorithm, called Uinta, to adaptively construct a structured P2P overlay network. Uinta not only takes advantages of physical characteristics of the network, but also places data belonging to the same semantic into a cluster and employs a class cache scheme to reduce the lookup routing latency. Simulations make some comparisons between Chord and our Uinta algorithm all running on the GT-ITM transit stub topology. The results show Uinta routing algorithm significantly improves P2P system lookup performance.