Design And Implementation Of Network Attack Situation Assessment Technology Based On Spark

With the development of communication technology China has entered the era of mobile Internet.There are more and more Internet users,the amount of data generated is increasing,and the network security problem is becoming more and more serious.Cyber attack situation assessment starts from the whole network,analyzes certain indicators in the network attack,judges the type of attack on the network,and designs corresponding indicators and algorithms based on the impact value of different attacks on the network.This project intends to implement a Spark-based network attack situation assessment platform to evaluate the network attack situation in a big data environment.The specific work is as follows:Based on the recurrent neural network and random forest algorithm,an RNN-RF attack detection model is constructed to evaluate the network attack situation.The recurrent neural network is responsible for dimensionality reduction of data features,and the random forest model is responsible for categorizing the dimensionality reduction data.The attack situation assessment algorithm combines the attack detection results of the RNN-RF model,analyzes the impact factors of the attack,calculates the network situation value,and combines the national network situation assessment standards to realize the assessment of the network attack situation.Experiments have proved that the RNN-RF model of this subject has better evaluation performance than other attack detection models.The RNN-RF model is simpler than the traditional mathematical model method and knowledge reasoning method,has a better evaluation effect on variant attacks,and can meet the needs of attack situation assessment in a big data environment.At the same time,this subject designs and implements a pluggable operator platform.Based on Spark's API and user requirements,this subject has designed and implemented a variety of pluggable operators.Specifically,it includes heterogeneous data input operators,row conversion operators,double set conversion operators,Key-Value conversion operators,SQL processing operators,action operators,and data output operators.The pluggable operator platform helps business personnel customize data processing and construct evaluation data,reducing the learning cost and use cost of business personnel using big data technology for data cleaning.Business personnel process the collected TCP data through the platform to convert it into the data form required by the algorithm,and quickly access it into the situation assessment model.The platform can easily utilize Spark's stream processing and batch processing capabilities to meet the needs of real-time evaluation and large-scale data evaluation.