【作者】 冷宁；

【导师】 周健；

【作者基本信息】 合肥工业大学 ， 计算机软件与理论， 2004， 硕士

【摘要】 网络规模和应用范围的迅速扩大，导致通过网络传播的病毒的数量越来越多，危害越来越大。在病毒尚未对网络使用造成严重危害前，采取有效的防护手段阻止病毒的传播成为一个研究的课题。 论文通过定期采集网络关键设备，包括汇聚层和核心交换层的流量情况，处理后以结构化、层次型的方式保存起来，一方面提供了图形化的用户接口，将收集的各种历史数据和实时数据在其中以图形的方式直观地显示出来，另一方面通过对历史流量数据的统计分析，从而对网络流量进行建模，短时间预测网络流量情况，实现对网络流量异常情况的观察分析，利用有充分响应能力的网络管理系统和网络安全设备，构成响应和预警的综合安全系统。 本系统在采用大型关系数据库对流量数据进行存储和管理的同时，采用Java及其分布式计算技术，很好的实现了逻辑业务层与数据层的分离，从而有较好的扩展性和可移植性。更多还原

【Abstract】 With rapid development of Internet and enlargement of range of network application, the quantity of virus which spread by network is getting more and more, its endangerment is getting greater and greater. Before the virus endangers our network seriously, we can take effective measures to prevent the spread of virus. More and more people are engaged in this research.This paper tries to gathering the traffic flow data of key equipment including influx layer and core exchange layer. This data will be saved structurally and gradationally. The system can provide graphical user interface to show the historical data and real-time data intuitively by graphics mode. On the other hand, by the statistical analysis of historical flow data, the system can model the network traffic flows, and forecast short-term network flows. After that the system can analyse the network traffic abnormity. By use of the netwok management system and network security facility, we can setup a security infrastructure to monitor flows in real time, predicts the future flows, alarm some abnormal flows, respond to the abnormity automatically and immediately.The system uses relational database management system for storage and .management of network flow data. By use of Java technology and its distributed computing environment, the system separates business layer from data layer commendably. Thus it has preferable extensibility and portability.更多还原