摘要
隐马尔可夫模型(HMM)是网络安全态势评估模型中应用最广泛的评估模型,其参数的确定直接影响了评估结果的客观性。针对这一问题,在粒子群算法(PSO)的基础上,提出自适应聚群粒子群算法(ASPSO)。首先通过人工鱼群在全局搜索方面的优势对PSO进行改进,然后对PSO算法中的惯性权值和学习因子进行改进,以提高HMM参数寻优准确率。最后以DARPA2000数据集中的LLDoS1.0的DDoS攻击场景进行模拟攻击,结果表明,改进算法在迭代次数方面要明显优于传统的PSO-HMM算法,且可真实模拟DDoS攻击场景,与云网络实际受到攻击时一致。
The Hidden Markov Model(HMM)is the most widely used evaluation model in the network security situation evaluation model,and the determination of its parameters directly affects the objectivity of the evaluation results.For this problem,the adaptive cluster particle group algorithm(ASPSO)is proposed based on the particle group algorithm(PSO).The PSO is first improved through the advantages of artificial fish groups in global search,and then the inertia weights and learning factors in the PSO algorithm are improved to improve the optimization accuracy of HMM parameters.Finally,the DDoS attack is the DDoS attack scene of LLDoS1.0 in the DARPA2000 dataset,which shows that the improved algorithm is significantly better than the traditional PSO-HMM algorithm in terms of iterations,and can truly simulate the DDoS attack scene,consistent with when the cloud network is actually attacked.
作者
郭义
Guo Yi(Fuyang Industrial Economics School,Fuyang 236502,China)
出处
《粘接》
CAS
2021年第11期179-183,共5页
Adhesion