Efficient Ethereum traffic identification method combining active node library and machine learning

The invention provides an efficient Ethereum traffic identification method combining an active node library and machine learning, which is divided into four parts: the first part is the construction of the active node library; the second part is training of a recognition model, and the third part is that different machine learning algorithms are used for comparative analysis, and a model obtained after training of the machine learning algorithm most suitable for classification is selected as the recognition model; and the fourth part is Ethereum traffic identification, specifically, the traffic is divided into TCP and UDP traffic after being screened by an active node library, the TCP and UDP traffic are input into an identification model for identification, and node information in the Ethereum active node library is updated according to an identification result. According to the method, the Ethereum traffic existing in the current network can be effectively identified, and the monitoring effect accuracy reac