Online data extraction for large-scale agent-based simulations

Daniel Zehe, Vaisagh Viswanathan, Wentong Cai, Alois Knoll

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

5 Scopus citations

Abstract

Cloud-based simulation systems reduce the upfront hardware costs of running high-performance experiments and increases the ease with which simulation experiments can be repeated. The data being generated by simulations can be large. Commonly used data storage systems such as relational databases can handle large amounts of data, but the analysis is a challenging problem. Moreover, handling this amount of data in cloud services can be both expensive (bandwidth and storage costs) and time-consuming. However, a lot of the data that is generated by agent-based simulations does not contribute directly to the purpose of the experiment being conducted. We propose an extension to cloud-based simulation systems that rather than storing raw simulation output data, uses stream data processing to generate the result dataset while the simulation is running. This can then be used to store only the data required for later use, this saving both time and money.

Original languageEnglish
Title of host publicationSIGSIM-PADS 2016 - Proceedings of the 2016 Annual ACM Conference on Principles of Advanced Discrete Simulation
PublisherAssociation for Computing Machinery, Inc
Pages69-78
Number of pages10
ISBN (Electronic)9781450337427
DOIs
StatePublished - 15 May 2016
Event2016 Annual ACM Conference on Principles of Advanced Discrete Simulation, SIGSIM-PADS 2016 - Banff, Canada
Duration: 15 May 201618 May 2016

Publication series

NameSIGSIM-PADS 2016 - Proceedings of the 2016 Annual ACM Conference on Principles of Advanced Discrete Simulation

Conference

Conference2016 Annual ACM Conference on Principles of Advanced Discrete Simulation, SIGSIM-PADS 2016
Country/TerritoryCanada
CityBanff
Period15/05/1618/05/16

Keywords

  • Agent-based simulation
  • Cloud-based simulation
  • Online data extraction
  • Time-variant relational algebra

Fingerprint

Dive into the research topics of 'Online data extraction for large-scale agent-based simulations'. Together they form a unique fingerprint.

Cite this