Research On Network Anomaly Detection Algorithm Based On Cloud Computing

With the development of the network, network attacks is increasing every day. For the high speed network,Traditional anomaly detection system is difficult to achieve the measurement of the whole network,and the traditional detection algorithms are also difficult to deal with the large scale data.Research on distributed anomaly detection system will be the future direction of development in the field of anomaly detection.This thesis puts forward a kind of anomaly detection system for large-scale network based on the cloud computing which has large storage and computing power. The system can real time measure of the network traffic,and achieve real time processing of abnormal.The main research work and achievements include:Firstly,To analyzes the status and related technology of network security,Summaries the existing problems of the traditional intrusion detection system,and the development direction in the future.And then,The present situation and the technology of flow measurement were introduces in the second chapter, It is important to measure the traffic of global network, beacuse it is more effective to detected abnormal as the input parameters.so It was detailed description of the process of this relevant part in the third chapter.Finally,Through analysis the specific realization of the anomaly detection method based on histogram,The fourth chapter realized the histogram algorithm in Cloud computing platform.In the experimental part,firstly it provided a case about how to detailed architecture the deployment of the network traffic measurement.and in the last,Through the implementation,the performance of the system and the algorithm was proved to be feasible.it can be used in the future field of the anomaly detection.