网络入侵检测系统的负载均衡方案

摘要

Network Intrusion Detection Systems(NIDS) often has high loss rate because of the high speed of data flow in high-speed net. This performance is improved by load balancing scheme developed for parallel NIDS based on prediction. This scheme takes the load of each sensor as observing value through relevant measurement,the global predictive model of chaotic time series as predictive method,and it balances the load according to the predictive load. Results of the simulation verify that the proposed scheme can balance load effectively and reduce the packet loss rate.%在高速网络环境下,数据流的高速化使得网络入侵检测系统往往会出现严重的漏报率,针对此性能瓶颈,提出了一种基于预测的并行入侵检测系统的负载均衡方案.该方案主动测量各探测器的负载为预测依据,采用混沌时间序列的全域预测法为预测手段,利用预测的负载值为负载均衡的根据.通过仿真实验,证明了该方案的可行性及有效性,它能有效地均衡负载、减少系统的丢包率.