The Passive Intrusion Detection Techniques Using WIFI

Passive Intrusion Detection technology overcomes the problem that video and image detection technology is limited by various environmental conditions, and the problem of the necessity to carry a wireless transceiver device. It shows important value and practical significance in areas like the elderly care, security monitoring and energy efficiency. With the continuous development of WIFI technology, it is widely used in the house, office and public locations because of its fast transmission speed, low cost and ease of installation, etc. Introducing WIFI technology in passive intrusion detection greatly improves the effective use of existing wireless network infrastructure, and at the same time, solves the problem of traditional intrusion detection methods requiring of special hardware support and testing equipment staying idle for most of the time.WIFI-based Passive Intrusion Detection is mainly implemented through detecting signal strength in the existing WIFI environment. By analyzing and processing the received signal strength indicator, people can extract and classify information that can characterize intrusion, to realize passive intrusion detection and identification. The main contents of this dissertation include the following aspects:1. The dissertation makes detailed analysis of the present situation of the development of Passive Intrusion Detection technology home and abroad and amply describes the research significance of WIFI-based Passive Intrusion Detection.2. The dissertation make indoor environment simulation, collecting much signal strength data of typical intrusion behavior like standing and moving in the detection area. We established samples feature set for classification and identification by analysis and pretreatment of related data.3. In order to achieve an efficient intrusion detection model, this dissertation made experiments to analyze the effects on identification results under different kernel functions. With the feature selection function in WEKA, grid search and cross-validation method, we made selections on feature set, slack variables and kernel function parameters, further improving the accuracy of detection and identification, and verifying the feasibility and effectiveness of WIFI-based Passive Intrusion Detection. The experimental results show that the model has a good effect on intrusion detection and identification.