Method of network traffic classification using Naïve Bayes based on FPGA

摘要

In order to solve the problems in High-speed network that the general software methods of traffic classification cannot meet the requirements of real-time; a method of Naïve Bayes based on FPGA for network traffic classification was proposed. This method is using Naïve Bayes based on FPGA for the network traffic classification, whose classification decisions can be reconfigured on the basis of classification results and network environment. In this way, the classification accuracy is guaranteed. And besides, FPGA-based Naïve Bayes, because of the hardware acceleration, is more rapid than Naïve Bayes based on software. Simulation results show that the former is 443 times than the latter in the classification rate. It can be inferred that this method is effectively for classification of High-speed network traffic.