Author(s)

Jinning Guo, Jiayi Zhao, Yuxuan Peng, Hao Wu

Corresponding Author

Jinning Guo

ABSTRACT

With the rapid development of modern network technology, serious network intrusion problems have arisen and the development of network intrusion detection technology has been promoted indirectly. In recent years, IDS based detection system has been gradually applied to all walks of life, playing an important role in monitoring and prevention. The core technology of intrusion detection also changes with the development of related disciplines. Machine learning and deep learning gradually replace the traditional detection technology based on state transition and monitoring with its solid theoretical foundation and strong adaptability, and begin to provide network security services for the new network represented by the Internet of things, making the network intrusion detection intelligent and adaptive. However, the current network intrusion detection still has low accuracy. This paper introduces the principle and working mode of mainstream intrusion detection technology based on the status of network intrusion phenomenon and the current commonly used models, points out the key problems existing in the industry at present, and forecasts the solution of the problems and the future development of network intrusion detection technology.

KEYWORDS

Intrusion detection, Machine learning, Deep learning, Internet of things