Study and application of BP neural network in intrusion detection

Abstract

Network security is the foundation to safeguard work concerning network, but various events and means to threaten network security have emerged one after another. As an effective auxiliary means in traditional security technology, the intrusion detection system has become a research hotspot in the field of computer system and network security. This paper has studied the basic algorithm of BP neural network, and adopted the improved LM-BP algorithm to design the intrusion detection system. Through comparison and analysis we find that compared with traditional algorithm of BP neural network, the algorithm of BP neural network not only has higher execution efficiency, its false alarm rate and error rate of intrusion detection are also lower than other detection methods, which has important significance to safeguard network security. © 2013 Springer Science+Business Media.