基于动态转移概率的网络态势预测方法研究

Research on Method of Network Situation Prediction Based on Dynamic Transition Probability

网络态势预测是网络态势感知领域的重要组成部分。鉴于目前网络状态变化莫测,网络攻击形式层出不穷,而现有的网络态势预测手段具有实时性差或者计算量繁重的特点,难以对网络态势进行准确有效的预测。对此,文中通过分析不同形式网络攻击下的网络状态变化特点,引入网络波动率的概念并用其描述网络波动状况,结合马尔可夫链提出一种基于动态转移概率的网络态势预测方法。该方法通过实时计算影响力衰减周期来计算转移概率,根据实际网络波动情况自适应更新转移概率的计算方式,从而动态地对网络态势进行预测,得到更准确有效的预测效果。对基于动态转移概率的网络态势预测模型进行了形式化描述并进行了原型实验,实验结果显示,该网络态势预测方法具有较高的准确性和可行性。

Network situation prediction is an important part of network situation awareness.In view of the unpredictable changes of network state and the emergence of network attack forms,the existing network situation prediction methods have the characteristics of poor real-time performance or heavy computation,which makes it difficult to predict the network situation accurately and effectively.For this,we introduce the concept of network volatility to describe the network fluctuation by analyzing the characteristics of network state changes under different forms of network attacks.Combining with Markov chain,a network situation prediction method based on dynamic transition probability is proposed.This method calculates the transition probability by calculating the attenuation period of influence in real time,and adaptively updates the calculation method of the transition probability according to the actual network fluctuation,so as to dynamically predict the network situation and achieve more accurate and effective prediction effect.In this paper,the network situation prediction model based on dynamic transition probability is formally described and prototype experiments are carried out.The experiment shows that the network situation prediction method has high accuracy and feasibility.