SEMSim Cloud Service: Large-scale urban systems simulation in the cloud

Daniel Zehe, Alois Knoll, Wentong Cai, Heiko Aydt

Research output: Contribution to journalArticlepeer-review

43 Scopus citations

Abstract

Large-scale urban systems simulations are complex and with a large number of active simulation entities the computational workload is extensive. Workstation computers have only limited capabilities of delivering results for large-scale simulations. This leads to the problem that many researchers and engineers have to either reduce the scope of their experiments or fail to execute as many experiments as they would like in a given time frame. The use of high-performance computing (HPC) infrastructure offers a solution to the problem. Users of such simulations are often domain experts with no or little experience with HPC environments. In addition users do not necessarily have access to an HPC. In this paper we propose an architecture for a cloud-based urban systems simulation platform which specifically aims at making large-scale simulations available to typical users. The proposed architecture also addresses the issue of data confidentiality. In addition we describe the Scalable Electro-Mobility Simulation (SEMSim) Cloud Service that implements the proposed architecture.

Original languageEnglish
Pages (from-to)157-171
Number of pages15
JournalSimulation Modelling Practice and Theory
Volume58
DOIs
StatePublished - Nov 2015

Keywords

  • Agent-based simulation
  • Cloud-based simulation
  • High performance computing
  • Performance evaluation
  • Simulation
  • Traffic simulation
  • Urban systems simulations

Fingerprint

Dive into the research topics of 'SEMSim Cloud Service: Large-scale urban systems simulation in the cloud'. Together they form a unique fingerprint.

Cite this