Cloud-based computer simulation: Towards planting existing simulation software into the cloud

详细信息	查看全文 | 推荐本文 |

• 作者：Xiaocheng Liu^a ; ^b ; ^{nudt200203012007xcl@gmail.com} ; Qiang He^a ; Xiaogang Qiu^a ; Bin Chen^a ; Kedi Huang^a
• 关键词：Cloud computing ; M&S ; Modeling as a service ; Simulation as a service ; Parallel Discrete Event Simulation
• 刊名：Simulation Modelling Practice & Theory
• 出版年：2012
• 期刊代码：80_1569190x
• 类别：mt
• 出版时间：August, 2012
• 卷：26
• 期：Complete
• 页码：135-150
• 文件大小：1501 K

摘要

Cloud computing provides various IT resources to users transparently in an easy-to-use, on-demand, cheap and pay-as-you-go manner. It frees users of burdens associated with managing computing resources and facilities, and reduces or even eliminates the capital outlays in hardware. These advantages are helpful to make simulation techniques become more accessible to users. Cloud-based computer Simulation (CSim) is becoming more and more attractive to Modeling and Simulation (M&S) practitioners. The primary challenge associated with CSim is how to make it practical. There are only a few of CSim research work proposed in the literature. To our best knowledge, none of them has tackled this question in full scale. This paper presents our work on planting existing simulation software into the cloud. The needs and the architecture of CSim, the development of CSim services, a modified modeling method and a novel simulation execution mode are proposed. In particular, some guidelines of making effective use of computing resources are developed after extensive experimentation. These guidelines provide the scheduler in CSim with useful approaches.