基于互联网的震后虚假信息入侵实时检测系统设计

Design of a Real-Time Detection System for False Post-EarthquakeInformation Intrusion Based on the Internet

摘要:

虚假震后信息入侵严重干扰抗震救灾与灾后重建进程，为防止虚假震后信息在互联网中大肆传播，需对震后虚假信息进行精准识别，设计基于互联网的震后虚假信息入侵实时检测系统。互联网以Web服务器为载体传播信息，数据包捕获模块采用WinPcap技术，并从中获取有效数据包，在协议解析模块中完成数据包TCP/IP协议、CMP协议、UDP协议解码工作，解码后的数据输入到基于SVM的震后虚假信息检测模型中。SVM模型寻求最优超平面将震后信息分为虚假与真实两个类别，根据该结果完成震后虚假信息入侵实时检测。由实验结果可知该系统检测震后虚假信息入侵误报率低于3%，相比同类型系统具有接收数据能力强、效率高的优势，对精准检测震后虚假信息具有重要意义。

Abstract:

False post-earthquake information intrusion seriously interferes with earthquake relief and postdisaster reconstruction. To prevent the spread of false post-earthquake information on the Internet, this study designs a real-time detection system. Web servers are used as carriers to transmit information on the Internet. Thus, the WinPcap technology is used herein to capture valid data packets. The transmission control protocol/internet protocol, control message protocol, and user datagram protocol are decoded in the protocol analysis module. The decoded data are used inputs in the false post-earthquake information detection model based on support vector machine. The model seeks the optimal hyperplane to divide the post-earthquake information into false and true. Results indicate the completion of the real-time detection of false post-earthquake information intrusion. The false alarm rate of this system is less than 3%. Different from related systems, the proposed system can receive data with high efficiency and is of great importance in accurately detecting false post-earthquake information.