The current Internet provides no information on the fate of transmitted packets. As a result, when packets get lost or delayed, there is no clean way for the affected parties to localize the problem and fix it (if it is local), ask for compensation (if a service-level agreement has been violated), or simply learn from it (e.g., re-assess a peering agreement with an under-performing neighbor). Probing tools like traceroute can help localize network failures, however, they draw their conclusions based on the fate of probes, not actual traffic, which makes them susceptible to manipulation by transit networks.
The goal of this project is a way to change this lack of accountability in the Internet: a clean, yet practical solution that tells network entities what they need to know (who is responsible for losing or delaying their packets), but not what they shouldn’t (the internal structure and policies of other networks or ISPs). The key idea behind our work is that this information need not (and should not) be extricated by ad-hoc probing tools that treat the Internet as a black box and try to reverse-engineer its structure and failures. Rather, it should be provided by a cooperative, incentive-based framework, where networks provide verifiable information on their own performance and, in exchange, learn how their own traffic is being treated by their neighbors.
In our 2010 paper, we show how the Internet network layer can provide an accurate, lie-resistant, and lightweight mechanism that gives information on the fate of transported traffic. It presents Network Confessional, a system and protocol that enables network domains to export information on their loss and delay performance. This system guarantees accurate and trustworthy information while minimizing in-network monitoring state, in particular, it requires no per-packet, per-flow, or per-path monitoring state inside the network.