Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
153 tokens/sec
GPT-4o
7 tokens/sec
Gemini 2.5 Pro Pro
45 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
38 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Dovetail: Stronger Anonymity in Next-Generation Internet Routing (1405.0351v2)

Published 2 May 2014 in cs.CR

Abstract: Current low-latency anonymity systems use complex overlay networks to conceal a user's IP address, introducing significant latency and network efficiency penalties compared to normal Internet usage. Rather than obfuscating network identity through higher level protocols, we propose a more direct solution: a routing protocol that allows communication without exposing network identity, providing a strong foundation for Internet privacy, while allowing identity to be defined in those higher level protocols where it adds value. Given current research initiatives advocating "clean slate" Internet designs, an opportunity exists to design an internetwork layer routing protocol that decouples identity from network location and thereby simplifies the anonymity problem. Recently, Hsiao et al. proposed such a protocol (LAP), but it does not protect the user against a local eavesdropper or an untrusted ISP, which will not be acceptable for many users. Thus, we propose Dovetail, a next-generation Internet routing protocol that provides anonymity against an active attacker located at any single point within the network, including the user's ISP. A major design challenge is to provide this protection without including an application-layer proxy in data transmission. We address this challenge in path construction by using a matchmaker node (an end host) to overlap two path segments at a dovetail node (a router). The dovetail then trims away part of the path so that data transmission bypasses the matchmaker. Additional design features include the choice of many different paths through the network and the joining of path segments without requiring a trusted third party. We develop a systematic mechanism to measure the topological anonymity of our designs, and we demonstrate the privacy and efficiency of our proposal by simulation, using a model of the complete Internet at the AS-level.

Citations (36)

Summary

We haven't generated a summary for this paper yet.