Mitigating Latency in Chord-Based Routing under IPv6

As the transition to IPv6 enables a massive expansion of the peer-to-peer (P2P) network landscape, traditional Chord-based routing protocols face significant performance bottlenecks due to the lack of awareness of the underlying physical network topology. In this paper, we propose eChord, a topology-aware routing system designed to mitigate latency in large-scale IPv6 environments by exploiting network locality. The key of eChord is a bifurcated routing architecture that maintains dual routing states: a localFinger table for intradomain routing within autonomous systems (AS) and a globalFinger table for inter-domain connectivity. By optimizing the identifier space and introducing a locality factor ρ to prioritize local lookups, eChord can effectively reduce the reliance on high-latency backbone links.We then perform amulti-dimensional efficiency analysis of the eChord system, in terms of expected hop number, end-to-end latency, and state maintenance overhead. Numerical simulations are finally provided to demonstrate that the proposed eChord system can significantly outperform standard Chord in various network scales. In particular, in a network of 106 nodes distributed across 500 ASs with a locality factor of ρ = 0.8, the proposed eChord system reduces the average routing latency by approximately 75% compared to the traditional, locality-agnostic Chord protocol.