一种基于遗传算法思想的流量监测系统的设计与实现

Design and Implementation of a Dataflow Monitoring System Based on the Genetic Algorithm Thought

摘要: 流量监测在网络管理中的承担着极其重要的角色，既可以了解当前网络的运行状态，判断其性能和网络故障，并能够反映安全时间发生与否，帮助网络安全的加强。本文所设计与实现的流量监测系统采用数据采集层、前置处理层、交互分析层的三层系统架构，并且为了高效地上传前置采集数据库信息到基础数据库中，在前置处理层中的程序中采用了多线程并行上传，并运用了遗传算法的思想为每条服务线程分配近乎相同的数据记录，从而保证数据上传在时间上的一致性。该系统实现了对大型城市办公网络的流量信息的实时监控、统计分析功能，并能以报表形式展示流量数据结果以作为网络管理和网络安全策略制定的依据。

Abstract: Traffic monitoring plays an extremely important role in network management, which can understand the current network running status, judge its performance and network failures, and also can reflect if it’s on the safety time or not, then help to strengthen the network security. Dataflow monitoring is an effective method to know the network running status, performance, some troubleshooting and enhance the network security. In this paper, the monitoring system is designed with data acquisition layer, pre-processing layer, and interaction analysis layer. To upload the local database information to the central database efficiently, the system uses multi-thread upload the record in the pre-processing program of the pre-processing layer and also uses the idea of the genetic algorithm to assign almost the same data records for each service thread to ensure the time consistency of data collection. The system achieves functions of real-time monitoring, statistics analyzing of the flow information for office network in one large city, and shows flow data results in report form as a basis for strategy formulation of network management and security.