Sim2HW: Modeling Latency Offset Between Network Simulations and Hardware Measurements

Johannes Späth, Max Helm, Benedikt Jaeger, Georg Carle

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Network modeling often relies on simulation tools due to their flexibility and cost-effectiveness. However, in many cases, those tools can only cover some aspects of real-world networks accurately. Measurements on hardware testbeds are more accurate but require more resources and configuration and are thus frequently impractical for real-world networks. Graph Neural Networks (GNNs) are a promising machine learning approach proven to be especially useful for learning the properties of computer networks. In this paper, we present a GNN-based approach that uses simulation data as an additional input to predict latency values measured on real hardware. We train our model with an existing dataset from a hardware testbed and show that it can predict the latency distribution in unseen topologies with a MAPE of 27.2 % and an MdAPE of 19.8 %.

Original languageEnglish
Title of host publicationGNNet 2024 - Proceedings of the 3rd GNNet Workshop on Graph Neural Networking Workshop, Co-Located with
Subtitle of host publicationCoNEXT 2024
PublisherAssociation for Computing Machinery, Inc
Pages20-26
Number of pages7
ISBN (Electronic)9798400712548
DOIs
StatePublished - 9 Dec 2024
Event3rd International Workshop on Graph Neural Networking, GNNet 2024, co-located with ACM CoNEXT 2024 - Los Angeles, United States
Duration: 9 Dec 202412 Dec 2024

Publication series

NameGNNet 2024 - Proceedings of the 3rd GNNet Workshop on Graph Neural Networking Workshop, Co-Located with: CoNEXT 2024

Conference

Conference3rd International Workshop on Graph Neural Networking, GNNet 2024, co-located with ACM CoNEXT 2024
Country/TerritoryUnited States
CityLos Angeles
Period9/12/2412/12/24

Keywords

  • Graph Neural Network
  • Hardware Measurement
  • Latency Model
  • Network Simulation

Fingerprint

Dive into the research topics of 'Sim2HW: Modeling Latency Offset Between Network Simulations and Hardware Measurements'. Together they form a unique fingerprint.

Cite this