A New Metric for Comparison of P2P Search Algorithms
Fatemeh Sharifkhani, Mohammad Reza Pakravan · 2012
Success rate, query hit rate and speed of search are some common metrics for evaluating p2p search algorithms. Yet, these metrics do not consider properties of discovered sources and their impact on download performance. A requesting peer is interested in stable serving peers with high upload bandwidth to minimize its download time and therefore, this is the superiority of a search algorithm to find sources with higher upload bandwidth in the p2p heterogeneous environment. In this paper, a new bio-inspired metric for comparing p2p search algorithms is introduced which takes the upload bandwidth of supplying nodes into account as a factor influencing quality of a search algorithm and user satisfaction. This metric is inspired from efficiency metric in foraging theory which assumes that animals search for nutrients in a way that maximizes their intake energy per spent foraging time.