Method and device for detecting network attack

The invention discloses a method and a device for detecting network attacks. The method comprises the following steps: receiving a network message based on a polling mode through a data plane development suite in a target chip; stripping data of the network message through a protocol stack, and inputting the data into a neural network model deployed in the target chip; and monitoring the data through the neural network model to detect a network attack, wherein the neural network model is a convolutional neural network-gated loop unit hierarchical neural network (RKNN) model. According to the invention, the traditional passive detection is changed into active detection, and the network attack to occur is pre-judged by using the artificial intelligence technology, so that the real-time performance and the accuracy of network attack detection are improved. 本发明公开了检测网络攻击的方法和装置。该方法包括：通过目标芯片中的数据平面开发套件基于轮询模式接收网络报文；通过协议栈剥离出所述网络报文的数据，并将所述数据输入至所述目标芯片中部署的神经网络模型；以及通过所述神经网络模型监测所述数据以检测网络攻击，所述神经网络模型为卷积神经网络-门控循环单元分层神经网络RKNN模型