一种基于马尔可夫的负载均衡方法

A Balancing Algorithm of Markov-based State

对于采用并行的高速网络入侵检测系统,负载均衡能力是一个重要的性能指标。多数的负载均衡算法都是根据检测引擎的负载情况来动态地分配数据流。提出了一种基于马尔可夫的负载均衡方法,根据流量分配的历史序列,来决定当前数据流的分配。同时结合当前系统的负载情况,来实现数据流的合理分配。

For the parallel high-speed network intrusion detection system,load balance ability is an important performance indicator.Most of the load balancing algorithms is based on the load of the detection engine,to the distribution of the dynamic data flow.This paper proposes a method called a balancing algorithm of markov-based state,and it is according to the flow rate distribution in the sequence of history to determine current data flow distribution,and this method is markov forecast method.Data flow distribution is realized by probability of every detective node with consideration of both markov state sequence and load conditions.