WS-PGRADE workflows for cloud-based distributed simulation

Abstract

Modeling and Simulation (M&S) is used for systems analysis and decision making in existing or new systems. Modeling large and complex organizations produces large-scale simulations that are difficult or impossible to run on a single computer. Such experiment execution requires high computation. Distributed Simulation (DS) allows modeling of large systems as smaller submodels that execute on different nodes of a computer network and interoperate with each other in order to compose larger systems. Furthermore, cloud computing offers on-demand access to multiple compute resources. Consequently, being able to run DS on cloud resources allows for more experimentation with large-scale simulations in a cost-effective way. However, deploying DS and in fact Cloud-based DS presents significant technical challenges. This paper proposes a framework for deploying DS on the cloud in a transparent manner using the CloudSME Simulation Platform based on WS-PGRADE workflows. A healthcare case study is used to demonstrate our approach.