2012 | OriginalPaper | Buchkapitel
Virtual Machine Task Allocation for HLA Simulation System on Cloud Simulation Platform
verfasst von : Shaoyun Zhang, Zhengfu Tang, Xiao Song, Zhiyun Ren, Huijing Meng
Erschienen in: AsiaSim 2012
Verlag: Springer Berlin Heidelberg
Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.
Wählen Sie Textabschnitte aus um mit Künstlicher Intelligenz passenden Patente zu finden. powered by
Markieren Sie Textabschnitte, um KI-gestützt weitere passende Inhalte zu finden. powered by
A new yet promising technology, Cloud computing, can benefit large-scale simulations by providing on-demand, everywhere simulation services to users. In order to enable multi-task and multi-user simulation tasks with Cloud computing, Cloud Simulation Platform (CSP) is proposed and developed. To promote the running efficiency of HLA systems on CSP, this paper proposes an approach addressing the Virtual Machine task allocation problem, which is divided into two levels of task allocation steps. The first-level uses a heuristic algorithm to optimize the mapping from federates (of HLA system) to virtual machines (of CSP) and aims to achieve load balance on virtual machines in CSP. The second-level dispatches the subtasks of federate to the cores of virtual machines to minimize the makespan (schedule length) of the federate which uses a DAG based list scheduling algorithm: the EST (Earliest-Start-Time) algorithm. Experiments show that the two-level task allocation strategy effectively improves the running efficiency of HLA system on CSP.