Authors: Francisco Pereira, Justine Sherry, Pratiksha Thaker
Organizations: INESC-ID, Instituto Superior Técnico, University of Lisbon, Lisbon, Portugal · Carnegie Mellon University, Pittsburgh, USA · Databricks, New York, USA
In 2003, Clark et al. proposed the "Knowledge Plane", a layer of intelligent agents that share rich information about network state in order to collaboratively manage the Internet across administrative boundaries. The technology of the day lacked the two capabilities it demanded most: reasoning and unstructured communication. Thus, automated network management has remained largely confined to single domains, while interdomain problems are still resolved by human operators on the phone. We argue that agentic large language models can finally unlock the potential of the Knowledge Plane, but that doing so will require solving new research problems that arise particularly in this decentralized, multi-agent, coopetitive setting that have received limited attention in the literature thus far.
Figures & tables
Figure 1. Simulated network for fault diagnosis wherein a student tries to access www.acm.org .
Figure 2. Topology of the Celer Bridge experiment. Arrows denote provider/client relationship, and solid lines peer/peer.
Figure 3. ISP routing decisions between Cheap and Expensive when running with Opus or GPT. The agent reasons about the 90th percentile billing on the Expensive option, rerouting traffic spikes to it without getting billed.