The Case for Hop-by-Hop Traffic Engineering

TL;DR

This paper advocates for Hop-by-Hop traffic engineering, which splits traffic at each router, offering a simpler, more efficient alternative to source-based routing like MPLS, without needing bandwidth reservations or demand predictions.

Contribution

It introduces a novel Hop-by-Hop traffic engineering approach, demonstrating its effectiveness and simplicity compared to traditional source routing methods.

Findings

HBH traffic engineering achieves near-optimal network resource utilization.

The restricted path choice impacts performance more than distributed decision-making.

Prototype results show performance close to the theoretical optimum.

Abstract

State-of-the-art Internet traffic engineering uses source-based explicit routing via MPLS or Segment Routing. Though widely adopted in practice, source routing can face certain inefficiencies and operational issues, caused by its use of bandwidth reservations. In this work, we make the case for Hop-by-Hop (HBH) Traffic Engineering: splitting traffic among nexthops at every router, rather than splitting traffic among paths only at edge routers. We show that HBH traffic engineering can achieve the original goals of MPLS (i.e., efficient use of network resources), with a much simpler design that does not need bandwidth reservations or predictions of traffic demand. We implement a prototype in the ns-3 network simulator, to investigate the cost imposed by 1) the restricted path choice of loop-free HBH multipath routing, and 2) the distributed decisions of each router, based on its local network view. We show that the former is more important than the latter, but that, other than a few outliers, our design shows a performance (= aggregate user utility) close to the theoretical optimum.