Resilience of dynamic overlays through local interactions
Proceedings of the 22nd international conference on World Wide Web companion
| Hi-index | 0.00 |
In a peer-to-peer (P2P) network, where nodes can freely join and leave at any time, the number of participating nodes tends to vary over time. Furthermore, in an unstructured P2P network, where any node can freely select direct linkages with any other node or nodes, this free selection/linkage process determines the network topology. A recent study that examined unstructured real-world P2P networks revealed that the observed topology of such networks will only follow a power-law degree distribution in a cumulative manner when the observation time period is long enough to allow numerous nodes to join and leave the network. Thus, to ensure the accuracy of simulation-based evaluations in unstructured P2P network research, this property, observed in real-world P2P networks, should also be reflected in topology generation for simulation. Accordingly, in this study, we propose a topology generation method in which the nodes participating in the network vary over time, a newly participating node randomly selects other nodes for linkage, and the resulting P2P network topology only follows a power-law degree distribution when the topology information collection time exceeds a certain duration. The proposed method is validated by simulation results.