CAJ | 학술논문

为了实现对现网大数据网络流量的实时、有效检测，提出了一种基于云计算的网络流量检测方案。该方案充分利用Hadoop平台Map/Reduce编程模型在海量数据处理方面的优势，采用分层化的设计思想，克服了传统检测方案在海量数据应用环境中效率低下、可扩展性与安全性不足的缺点。重庆移动DPI平台应用表明，该方案较为有效，流量检测效果良好，在大数据处理时效率较普通分布式处理有明显提高。
위료실현대현망대수거망락류량적실시、유효검측，제출료일충기우운계산적망락류량검측방안。해방안충분이용Hadoop평태Map/Reduce편정모형재해량수거처리방면적우세，채용분층화적설계사상，극복료전통검측방안재해량수거응용배경중효솔저하、가확전성여안전성불족적결점。중경이동DPI평태응용표명，해방안교위유효，류량검측효과량호，재대수거처리시효솔교보통분포식처리유명현제고。
In order to detect the network traffic in real time and effectively, a traffic detection scheme based on cloud computing is proposed. The scheme makes full use of the advantages of Hadoop platform's Map/Reduce programming model and adopts the thought of hierarchical design to avoid the inefficiency, the de-fects of scalability and security of traditional detection schemes in big data application environment. The ap-plication of the proposed scheme in DPI platform of Chongqing Mobile Communication Company has proved its effictiveness. The detection result is satisfying and the data processing speed has been improved obvi-ously compared with the common distributed processing.