基于改进隐马尔可夫模型的云网络安全研究

Research on Cloud Network Security based on Improved Hidden Markov Model

针对隐马尔可夫(HMM)模型参数选择存在很大的主观性问题,提出一种基于自适应聚群粒子群算法(ASPSO)优化HMM的云网络安全态势评估方法。首先通过人工鱼群提高PSO算法的全局搜索性能,同时改进PSO的惯性权值和学习因子进,以提高HMM参数寻优准确率;然后以构建的最优HMM模型,构建云网络安全态势评估模型;最后模拟DDoS攻击场景,对上述评估模型进行验证。结果表明,改进算法在HMM参数寻优方面,只需迭代160次左右,而传统的PSO优化寻优要迭代430次。同时在真实模拟DDoS攻击场景时,与云网络实际受到攻击时大致相同,且在不同阶段表现出不同的态势值。由此看出,该改进模型可有效预测网络的安全。

Aiming at the subjective problem of hidden Markov(HMM)model parameter selection,a cloud network security situation assessment method based on adaptive clustering particle swarm optimization(aspso)HMM is proposed.Firstly,the global search performance of PSO algorithm is improved by artificial fish swarm,and the inertia weight and learning factor of PSO are improved to improve the accuracy of HMM parameter optimization.Then,a cloud network security situation assessment model is constructed based on the constructed optimal HMM model.Finally,the DDoS attack scenario is simulated to verify the above evaluation model.The results show that the improved algorithm only needs about 160 iterations in HMM parameter optimization,while the traditional PSO optimization needs 430 iterations.At the same time,in the real simulation of DDoS attack scenario,it is roughly the same as the situation when cloud network is actually attacked,and shows different situation values in different stages.It can be seen that the improved model can effectively predict the security of the network.