Latency-Aware Inter-domain Routing
Shihan Lin, Yi Zhou, Xiao Zhang, Todd Arnold, Ramesh Govindan, Xiaowei Yang
TL;DR
The paper tackles the persistent problem of BGP-induced latency inflation by proposing two lightweight, incremental methods: latency-proportional AS path prepending and local preference neutralization. Through CAIDA ITDK-based router-level simulations, the authors show that these mechanisms can substantially reduce end-to-end latency—notably a >31% reduction at the 90th percentile—while incurring roughly half the message overhead of an ideal minimum-latency routing protocol. The work analyzes practical deployment considerations, tradeoffs in quantization of latency and routing dynamics, and outlines a path toward real-world validation and integration with existing business models. Overall, the study demonstrates that modest, deployable changes to BGP can meaningfully mitigate tail latency for latency-sensitive services in inter-domain routing.
Abstract
Despite efforts from cloud and content providers to lower latency to acceptable levels for current and future services (e.g., augmented reality or cloud gaming), there are still opportunities for improvement. A major reason that traffic engineering efforts are challenged to lower latency is that the Internet's inter-domain routing protocol, the Border Gateway Protocol, is oblivious to any performance metric, and circuitous routing is still pervasive. In this work, we propose two implementation modifications that networks can leverage to make BGP latency-aware and reduce excessive latency inflation. These proposals, latency-proportional AS prepending and local preference neutralization, show promise towards providing a method for propagating abstract latency information with a reasonable increase in routing overhead.