Abstract: Research on DDoS attack is one of the hotspots in network security field.Forecasting the probability of DDoS attack is highly significative of defending and responding DDoS attack at its early stage.In this paper，a forewarning model for DDoS attack based on Probable Finite State Automation（PFSA） is proposed.Moreover，a method to calculate the probability of making attack and the degree of expectation threat is presented.Experiment indicates that the method proposed in this paper can calculate probability and expectation damage of DDoS attack effectively，which achieves the goal of early defence.

Key words: Distributed Denial of Service（DDoS） attack, forewarning model, Probable Finite State Automation（PFSA）, feature extraction

摘要： DDoS攻击的研究是网络安全研究领域的一个研究热点。预测DDoS攻击发生的概率，对于早期防御和响应DDoS攻击具有重要意义。提出了一种基于概率有限状态自动机的DDoS攻击预警模型，给出了一种计算攻击成功概率和攻击期望威胁度的方法。实验表明，方法能有效地计算出DDoS攻击成功概率和期望威胁度，实现早期防御。

关键词: 分布式拒绝服务（DDoS）攻击, 预警模型, 概率有限状态自动机, 特征提取