Design And Implementation Of Network Anomaly Detection System Based On Weighted Naive Bayes

Since the emergence of computer network technology in the 1970 s, until now the network technology has deep into every aspect of our life, computer networks have become an indispensable part of our life. But while enjoy the convenience of network, network attacks incidents such as viruses, hobbyhorse, hacker, network crime also bring great threat to our network security, especially in today's social the ongoing convergence of network and reality in life such as currency, trade and information exchange activities, this situation increased the network real interests of the world, more and more hosting such a business network has become the focus of network attack to some extent.In view of the ongoing network crime, we must first do nip in the bud. In respect of network crime prevention the common method is to develop the network anomaly detection system, in front of the network attack caused serious harm effectively detect attacks and make the relevant preventive measures. Based on the relevant technology and theoretical knowledge, this paper developed a network anomaly detection system based on weighted naive bayes. We can take use of the system to analysis and detect the network attack, and after the system test found that the system can effectively detect the network attack. The main work of this paper is shown below:(1)On the basis of consulting a large number of related references at home and abroad, analysis the current background and development status quo of network anomaly detection system, then further studied the theoretical basis and the relevant technology of the network anomaly detection, finally cleared the design goal of the system. Then analysis the system requirements which mainly include the functional requirements and non-functional requirements.(2)Based on the system demand analysis, proposed an improved weighted naive bayes algorithm for network anomaly detection, and designed a network anomaly detection system based on the algorithm. For in the design process, firstly we do a summary design of the system which include the technical architecture and function architecture, then determine three functional modules of the system: data collection module, data storage module and data analysis module and determine technology and methods for the each module and the work flow between modules.(3)Implement the system according to the system requirements analysis and system design. The system uses the C/S architecture, implement the related functions of network anomaly detection system based on weighted naive bayes through the programming. Then test the system in local area network in the laboratory environment. By the system test, we found that this system can accurately detect the common network attacks and other auxiliary functions, and has high detection accuracy and speed.