Traffic classification method and system based on FPGA random forest model

Abstract

The invention relates to a traffic classification method and system based on an FPGA random forest model. The method comprises the following steps: acquiring to-be-classified network data streams to be classified and a random forest model trained by software; controlling an FPGA to store the network data streams to be classified and loading the random forest model; extracting characteristic parameters of the network data streams to be classified; and classifying the network data streams to be classified according to the characteristic parameters and the random forest model. According to the traffic classification method and system based on an FPGA random forest model, random forest model parameters can be directly modified through software, data traffic classification is achieved through FPGA hardware, when the random forest model parameters need to be modified, the random forest model parameters can be directly modified through the software, the FPGA hardware does not need to be modified, the software serves as hardware drive of the FPGA, and therefore the classification speed of the network data traffic is increased.

Description

technical field [0001] The present invention relates to the field of traffic classification, in particular to a traffic classification method and system based on a Field Programmable Gate Array (Field Programmable GateArray, FPGA) random forest model. Background technique [0002] Traffic classification technology is the basis of network management and security. In the current network, the proportion of encrypted traffic continues to grow. There are a large number of encrypted traffic based on protocols such as SSL / TLS / VPN in the network. Network traffic classification also brings challenges, which brings new challenges to traffic collection and restoration. However, traditional technologies based on ports, protocols, and dpi engines can no longer meet the requirements for real-time traffic classification. That is, the existing traffic classification method has the problem of slow classification speed, and cannot realize real-time classification of network traffic. Conten...

AI Technical Summary

Problems solved by technology

In the current network, the proportion of encrypted traffic continues to grow. There are a large number of encrypted traffic based on protocols such as SSL / TLS / VPN in the network. Network traffic classification also brings challenges, which brings new challenges to traffic collection and restoration. However, traditional technologies based on ports, protocols, and dpi engines can no longer meet the needs of real-time traffic classification

That is, the existing traffic classification method has the problem of slow classification speed, and cannot realize real-time classification of network traffic.

Images