Exploiting autonomous system information in structured peer-to-peer networks

With the rise of peer-to-peer networks, two problems have become prominent: (1) significant network traffic among peers, to probe the latency among those peers to improve lookup performance; (2) the need for increased information flow across protocol layer boundaries, to allow for cross-layer adaptations. The particular work here focuses on improvements in structured peer-to-peer networks based on autonomous system information. We find that by using autonomous system information effectively we can achieve lookup performance approaching that based on proximity neighbor selection, but with much less network traffic. We also demonstrate improvements in replication in structured peer-to-peer networks using AS topology and scoping information. Finally, we review this approach in the context of network architecture