Research On Trace Back Technology Of Web Attack Based On Big Data

Web application is kind of used in various fields of network application widely,like network shopping,social networking,online banking,blogs and Web mail.With the development of Web application continuously,the security problem prominent increasingly and its harm is larger and larger.Web application consists of two parts:browser client program and server program.Web attack refers to the use of a loophole in the browser and server,through various illegal ways to obtaining the what they need.With the big data coming,vast amounts of data and the changing of data structure,the number of traditional network attack source is increasingly expanding,more complexity and correlation.The traditional Web attack detection technology already does not adapt to the current demand.On the basis of large data,linkage as the main line and realizing the trace back,this thesis focuses on Web attack source tracking.Based on BP algorithm to optimize packet marking scheme.Firstly,state this article research background and significance,summarizes the present situation of the research at home and abroad.Respectively from the principle of Web attack,attack types and the characteristics of Web attacks to analyze Web attack in big data environment.Study the existing race back technology,the advantages and disadvantages of various types of technology were compared,for the follow-up depth study trace back technology laid the foundation.Secondly,select the packet marking strategy for the study,and depth study the existing packet marking technology.Compare the probability of packet marking technology and the determine packet marking technology,and analyzes the basic principles of BP Algorithm,put forward a new packet marking technology based on BP Algorithm.Through the establishment of Web attack model,adjust the weights of neural network,improve marking scheme.Finally,on NS2 simulation platform,in IPv4 networks to attack path reconstruction,and analysis the simulation results and the performance of this trace back scheme.