With the continuous expansion of the data center network scale, changing network requirements, and increasing pressure on network bandwidth, the traditional network architecture can no longer meet people’s needs. The...With the continuous expansion of the data center network scale, changing network requirements, and increasing pressure on network bandwidth, the traditional network architecture can no longer meet people’s needs. The development of software defined networks has brought new opportunities and challenges to future networks. The data and control separation characteristics of SDN improve the performance of the entire network. Researchers have integrated SDN architecture into data centers to improve network resource utilization and performance. This paper first introduces the basic concepts of SDN and data center networks. Then it discusses SDN-based load balancing mechanisms for data centers from different perspectives. Finally, it summarizes and looks forward to the study on SDN-based load balancing mechanisms and its development trend.展开更多
Extensive investigation has been performed in location-centric or geocast routing protocols for reliable and efficient dissemination of information in Vehicular Adhoc Networks （VANETs）. Various location-centric rout...Extensive investigation has been performed in location-centric or geocast routing protocols for reliable and efficient dissemination of information in Vehicular Adhoc Networks （VANETs）. Various location-centric routing protocols have been suggested in literature for road safety ITS applications considering urban and highway traffic environment. This paper characterizes vehicular environments based on real traffic data and investigates the evolution of location-centric data dissemination. The current study is carded out with three main objectives： （i） to analyze the impact of dynamic traffic environment on the design of data dissemination techniques, （ii） to characterize location-centric data dissemination in terms of functional and qualitative behavior of protocols, properties, and strengths and weaknesses, and （iii） to find some future research directions in information dissemination based on location. Vehicular traffic environments have been classified into three categories based on physical characteristics such as speed, inter-vehicular distance, neighborhood stability, traffic volume, etc. Real traffic data is considered to analyze on-road traffic environments based on the measurement of physical parameters and weather conditions. Design issues are identified in incorporating physical parameters and weather conditions into data dissemination. Functional and qualitative characteristics of location-centric techniques are explored considering urban and highway environments. Comparative analysis of location-centric techniques is carded out for both urban and highway environments individually based on some unique and common characteristics of the environments. Finally, some future research directions are identified in the area based on the detailed investigation of traffic environments and location-centric data dissemination techniques.展开更多
深入分析交通事故数据可以为规避事故发生、降低事故严重程度提供重要理论依据,然而,在事故数据采集、传输、存储过程中往往会产生数据缺失,导致统计分析结果的准确性下降、模型的误判风险上升。本文以芝加哥2016—2021年的101452条追...深入分析交通事故数据可以为规避事故发生、降低事故严重程度提供重要理论依据,然而,在事故数据采集、传输、存储过程中往往会产生数据缺失,导致统计分析结果的准确性下降、模型的误判风险上升。本文以芝加哥2016—2021年的101452条追尾事故数据为研究对象,将原始数据按照7∶3随机分为训练集和测试集。在训练集数据上,利用生成式插补网络(Generative Adversarial Imputation Network,GAIN)实现对缺失数据的填补。为对比不同数据填补方法的效果,同时选择多重插补(Multiple Imputation by Chained Equations,MICE)算法、期望最大化(Expectation Maximization,EM)填充算法、缺失森林(MissForest)算法和K最近邻(K-Nearest Neighbor,KNN)算法对同一数据集进行数据填补,并基于填补前后变量方差变化比较不同填补算法对数据变异性的影响。在完成数据填补的基础上,构建LightGBM三分类事故严重程度影响因素分析模型。使用原始训练集数据,以及填补后的训练集数据分别训练模型,并使用未经填补的测试集数据检验模型预测效果。结果表明,经缺失值填补后,模型性能得到一定改善,使用GAIN填补数据集训练的模型,相较于原始数据训练的模型,准确率提高了6.84%,F1提高了4.61%,AUC(Area Under the Curve)提高了10.09%,且改善效果优于其他4种填补方法。展开更多
网络流量数据的高维复杂特性,使得生成对抗网络生成的网络流量数据质量较差。为了解决该问题,提出一种基于双生成器的条件映射生成对抗网络(a cGAN with projection discriminator based on double generators,PD-DcGAN)并将其应用于少...网络流量数据的高维复杂特性,使得生成对抗网络生成的网络流量数据质量较差。为了解决该问题,提出一种基于双生成器的条件映射生成对抗网络(a cGAN with projection discriminator based on double generators,PD-DcGAN)并将其应用于少数类流量增强。提出基于Gumbel-sigmoid分布的离散生成器,获得近似于离散数据的光滑可导分布生成离散特征,并将其与连续数据生成器并联运行,二者结果串联组合,获得数据整体分布情况;以内积形式融合条件信息和特征信息,克服传统方法出现假设空间增大的问题,缓解模型训练过程中的不稳定现象;在损失函数中引入梯度惩罚因子,将判别器梯度限定在一定范围内,有效缓解梯度爆炸。利用UNSW-NB15数据集,从生成样本质量和模型有效性两个角度检验模型性能。实验结果证明,与其他数据增强方法相比,PD-DcGAN在准确率、精确率、召回率和F1得分上分别平均提高2.72%、1.72%、1.87%和1.16%;与原始数据集相比,对难以检测的Analysis、Backdoors、Exploits、Shellcode和Worms等少数类流量检测性能提升明显,分别从不足1%分别提升至7.93%、6.53%、15.72%、14.02%和10.91%。展开更多
文摘With the continuous expansion of the data center network scale, changing network requirements, and increasing pressure on network bandwidth, the traditional network architecture can no longer meet people’s needs. The development of software defined networks has brought new opportunities and challenges to future networks. The data and control separation characteristics of SDN improve the performance of the entire network. Researchers have integrated SDN architecture into data centers to improve network resource utilization and performance. This paper first introduces the basic concepts of SDN and data center networks. Then it discusses SDN-based load balancing mechanisms for data centers from different perspectives. Finally, it summarizes and looks forward to the study on SDN-based load balancing mechanisms and its development trend.
文摘Extensive investigation has been performed in location-centric or geocast routing protocols for reliable and efficient dissemination of information in Vehicular Adhoc Networks （VANETs）. Various location-centric routing protocols have been suggested in literature for road safety ITS applications considering urban and highway traffic environment. This paper characterizes vehicular environments based on real traffic data and investigates the evolution of location-centric data dissemination. The current study is carded out with three main objectives： （i） to analyze the impact of dynamic traffic environment on the design of data dissemination techniques, （ii） to characterize location-centric data dissemination in terms of functional and qualitative behavior of protocols, properties, and strengths and weaknesses, and （iii） to find some future research directions in information dissemination based on location. Vehicular traffic environments have been classified into three categories based on physical characteristics such as speed, inter-vehicular distance, neighborhood stability, traffic volume, etc. Real traffic data is considered to analyze on-road traffic environments based on the measurement of physical parameters and weather conditions. Design issues are identified in incorporating physical parameters and weather conditions into data dissemination. Functional and qualitative characteristics of location-centric techniques are explored considering urban and highway environments. Comparative analysis of location-centric techniques is carded out for both urban and highway environments individually based on some unique and common characteristics of the environments. Finally, some future research directions are identified in the area based on the detailed investigation of traffic environments and location-centric data dissemination techniques.
文摘深入分析交通事故数据可以为规避事故发生、降低事故严重程度提供重要理论依据,然而,在事故数据采集、传输、存储过程中往往会产生数据缺失,导致统计分析结果的准确性下降、模型的误判风险上升。本文以芝加哥2016—2021年的101452条追尾事故数据为研究对象,将原始数据按照7∶3随机分为训练集和测试集。在训练集数据上,利用生成式插补网络(Generative Adversarial Imputation Network,GAIN)实现对缺失数据的填补。为对比不同数据填补方法的效果,同时选择多重插补(Multiple Imputation by Chained Equations,MICE)算法、期望最大化(Expectation Maximization,EM)填充算法、缺失森林(MissForest)算法和K最近邻(K-Nearest Neighbor,KNN)算法对同一数据集进行数据填补,并基于填补前后变量方差变化比较不同填补算法对数据变异性的影响。在完成数据填补的基础上,构建LightGBM三分类事故严重程度影响因素分析模型。使用原始训练集数据,以及填补后的训练集数据分别训练模型,并使用未经填补的测试集数据检验模型预测效果。结果表明,经缺失值填补后,模型性能得到一定改善,使用GAIN填补数据集训练的模型,相较于原始数据训练的模型,准确率提高了6.84%,F1提高了4.61%,AUC(Area Under the Curve)提高了10.09%,且改善效果优于其他4种填补方法。
文摘网络流量数据的高维复杂特性,使得生成对抗网络生成的网络流量数据质量较差。为了解决该问题,提出一种基于双生成器的条件映射生成对抗网络(a cGAN with projection discriminator based on double generators,PD-DcGAN)并将其应用于少数类流量增强。提出基于Gumbel-sigmoid分布的离散生成器,获得近似于离散数据的光滑可导分布生成离散特征,并将其与连续数据生成器并联运行,二者结果串联组合,获得数据整体分布情况;以内积形式融合条件信息和特征信息,克服传统方法出现假设空间增大的问题,缓解模型训练过程中的不稳定现象;在损失函数中引入梯度惩罚因子,将判别器梯度限定在一定范围内,有效缓解梯度爆炸。利用UNSW-NB15数据集,从生成样本质量和模型有效性两个角度检验模型性能。实验结果证明,与其他数据增强方法相比,PD-DcGAN在准确率、精确率、召回率和F1得分上分别平均提高2.72%、1.72%、1.87%和1.16%;与原始数据集相比,对难以检测的Analysis、Backdoors、Exploits、Shellcode和Worms等少数类流量检测性能提升明显,分别从不足1%分别提升至7.93%、6.53%、15.72%、14.02%和10.91%。
