网络实验平台与计算实验方法研究与展望  

Network Simulation and Computational Experiment Platforms:the Overview and Prospect

作　　者：杨林瑶 李玉珂[1,2,3] 韩双双 王晓[1,2,4] YANG Lin-Yao;LI Yu-Ke;HAN Shuang-Shuang;WANG Xiao(The State Key Laboratory for Management and Control of Complex Systems,Institute of Automation,Chinese Academy of Sciences,Beijing 100190,China;Qingdao Academy of Intelligent Industries,Qingdao Shandong 266000,China;Vehicle Intelligence Pioneers Inc.,Qingdao Shandong 266000,China;University of Chinese Academy of Sciences,Beijing 100080, China)

机构地区：[1]中国科学院自动化研究所复杂系统管理与控制国家重点实验室,北京100190 [2]青岛智能产业技术研究院,山东青岛266000 [3]中国科学院大学,北京100080 [4]青岛慧拓机器智能有限公司,山东青岛266000

出　　处：《指挥与控制学报》2018年第4期281-290,共10页Journal of Command and Control

基　　金：国家自然科学基金(61501461;61702519;61533019;91720000);北京市科学技术委员会项目人机意图智能融合的平行驾驶技术研究及应用验证(2181100008918007);英特尔智能网联汽车大学合作研究中心(ICRI-IACV)资助~~

摘　　要：主要介绍网络仿真技术的原理、特点及其通用的网络仿真平台,对不同的网络仿真软件进行了对比分析与评价.针对传统的仿真方法存在的弊端,结合网络发展的需求和趋势,介绍了平行网络的架构及其计算实验方法.该方法可以充分利用网络大数据,通过多种实验方法分析网络特征,拟合并指导真实的网络自适应优化.该方法具有很好的模拟准确度和虚实交互性,并且能够低成本地自动优化网络,以满足现实的需要.In this paper,we introduce the principle,the characteristics,and the main common platforms of network simulation,and some comparisons and evaluations are presented as well.To overcome the drawbacks of the traditional simulation methods according to the requirements and tendency of networks,we introduce a parallel network as well as its computational experiment platform.Through the continuous interaction and optimization between artificial network and physical network,the parallel network has the capability of improving the performance of the .overall network.Based on the complete artificial-network data and physical network data,the computational experiment method can learn and evaluate different operational schemes under various scenarios,and provide the optimal scheme at low cost.

关 键 词：网络仿真 仿真平台 平行网络 计算实验 虚实交互 网络优化 

分 类 号：TP393.0[自动化与计算机技术—计算机应用技术]