Study On Visual Analytics For Network Traffic Logs

With the rapid development of the Internet,Internet of things,cloud computing and other information technology,we have entered the era of big data.Big data analysis is one of the core contents of big data research.There are two ways to analysis the big data.The first one use the machines for analysis.These machines need be very powerful and the method of analysis is artificial intelligence.In the other one,people play the key roles in the analysis process by using the ability that human have but machines not,such as understanding a painting.The data visualization analysis belongs to the second category.This paper is using the data visualization analysis as a tool to find the interesting information that hide in the network log.This paper proposed a visualization solution to satisfy the network administrators' network monitoring requirements that the network administrators can analyze the network for discovering the abnormal events,user behaviors and the whole situation of network services.We proposed some visualization layouts to meet these requirements.According to the multi-view collaborative analysis and cases study,the effectiveness of the proposed solution is verified.The main work of this paper includes:(1)Designed several views to display the main information of network log.The main information of network log are source IP,destination IP,source port,destination port,timestamp and the size of the transmitted data.The connection view use the force-oriented model to calculate the position of each terminal node.(2)After the views are well designed,we use D3.js,Vue.js and some other open source JavaScript libraries to achieve these views.Moreover,the views are optimized according to the features of Canvas and SVG.(3)We utilize the ChinaVis2016 challenge data as case study.Firstly,we do some preprocessing on the data by some visualization method.Then,we use the preprocessed data to do the case study.In collaborative analysis of the data,we find the network running status,host behaviors and network service situation of the network log data.