针对固定移动融合(FMC)网络环境下的垃圾语音信息(Spam over Internet Telephony,SPIT)隐患,提出了一种多层次的检测方法。该方法首先深入剖析SPIT攻击独有的特点,增强检测的针对性。然后分多个层次进行检测,信令监控算法根据SPIT独有的特点来监控所有SPIT相关的原子信息;统计分析算法对原子信息进行统计,得到具有统计特性的更高语义级别的告警;关联分析算法对告警进行关联,识别出攻击的行为踪迹,重构攻击场景。仿真实验与分析证明了该方法的功能与性能满足SPIT检测的需求。 Aiming at the hidden trouble of Spam over Internet Telephony (SPIT) in fixed mobile convergence (FMC) network environment, a multi-level detection method is proposed. The method first analyzes the unique features of SPIT attack in depth and enhances the pertinence of the detection. And then detect it at multiple levels. The signaling monitoring algorithm monitors all the SPIT-related atomic information according to the SPIT’s unique features. The statistical analysis algorithm collects the atomic information to obtain higher semantic level alarms with statistical characteristics. Correlation analysis The algorithm correlates the alarms, identifies the traces of the attack, and reconstructs the attack scenario. The simulation experiment and analysis prove that the function and performance of the method meet the requirements of SPIT detection.