An Environmental Intrusion Detection Technology Based on WiFi

Abstract

The traditional intrusion detection technology has some shortcomings, such as high hardware requirements and harsh detection conditions etc. This paper proposes an environment intrusion detection technology based on WiFi, which uses the existing WiFi network to realize security monitoring function, covers a wide range and does not expose privacy. Firstly, the technology uses median filtering to denoise the subcarriers in the channel, and then using the self-organizing competitive neural network algorithm for fingerprint feature extraction and establish the intrusion signal. Finally, the statistical model of the nonlinear dependence between the intrusion and the fingerprint database is obtained by using the classification of normalized exponential function, to achieve the purpose of intrusion detection. The experimental results show that the recognition rate of this technology is improved by nearly 8% compared with the existing methods, reaching 98%, which has a good development prospect.