Peer-to-peer traffic recognition in cloud computing environment

Peer-to-peer traffic recognition in cloud computing environment

Tankai Gaozhonger Li Fengban

Because of the memory limit, the Peer-to-peer traffic recognition method can only deal with the small-scale data set, and the attribute characteristics of the recognition method based on naive Bayesian classification are artificial selection, so the recognition rate is limited and lacks objectivity. Based on the above problem analysis, this paper puts forward naive Bayesian classification algorithm in cloud computing environment and improves the attribute reduction algorithm in cloud computing environment, and combines these two algorithms to realize the fine granularity recognition of the encrypted peer-to-peer traffic. Experimental results show that this method can efficiently deal with large data set network traffic, and has a high rate of peer-to-peer traffic recognition, and the results are also objective.

Peer-to-peer traffic recognition in cloud computing environment