A P2P-based system for searching in metric spaces

In this paper, we elaborate on a scalable and distributed similarity search structure, that is the problem, which has previously been studied only for single computers. Our structure is scalable in that it distributes the data over more and more independent peer computers. It has no hot spot {--} all peers use as precise addressing scheme as possible and they all incrementally learn from misaddressing. Updates are performed locally and a node splitting never requires sending multiple messages to many peers. Experiments conducted on a prototype system are also reported.