改进的粒子群算法在入侵检测中的应用

Application of intrusion detection based on improved particle swarm optimization

为了提高入侵检测系统的检测率和降低系统的误检率,对基本的粒子群算法采用在粒子群初始化阶段,种群的离散度必须满足一定的要求才能开始迭代;在算法迭代过程中,惯性权重、加速系数的调整都与当前粒子群的离散度相关;当种群的离散度小于一定数值时,进行保优重初始化,同时采用适应度函数拉伸操作,重新迭代等几个方面的改进。经过KDD Cup 1999数据集的训练和检验数据的仿真测试,改进后的粒子群算法具有较高的检测正确率和较低的误检率,而且新算法收敛速度快,不易局部最优。

In order to improve the detection rate and reduce the false detection rate of intrusion detection systems , In the initialization phase of particle swarm, discrete degree of swarm must meet certain requirements before its iteration. In the process of iterative algorithm, the adjustment of inertia weight and acceleration coefficient was related to current discrete degree of particle swarm. When discrete degree was smaller than certain value, it should reinitialize in order to retain high quality, stretch fitness function and reiterate. The improved particle swarm optimization algorithm has higher detection accuracy and lower false detection rate, The new algorithm has the advantages of fast convergence and difficult to local optimum.