基于WIFI的被动入侵检测技术研究

发布时间：2018-03-07 16:21

本文选题：WIFI　切入点：被动入侵检测　出处：《合肥工业大学》2014年硕士论文　论文类型：学位论文

【摘要】：被动入侵检测技术克服了视频和图像检测技术受环境条件限制的问题,且检测目标无需携带无线收发装置,在老年人监护、安全监控和节能等领域有着非常重要的研究价值和现实意义。随着WIFI技术的不断发展,WIFI技术以其传输速度快、成本较低和易于安装等优点,在家庭、办公场所和公共地段得到广泛地应用。利用WIFI进行被动入侵检测,可以更加有效的使用现有无线网络基础设施,同时能解决传统入侵检测方法需要特殊硬件支持,以及检测设备大部分时间处于闲置状态的问题。本文所研究的基于WIFI的被动入侵检测方法,主要采用现有WIFI环境中的无线信号实现被动入侵检测。通过对接收信号强度(Received Signal Strength Indicator,简称RSSI)进行分析处理,提取能够表征入侵行为的特征,并对其进行分类识别,达到被动入侵检测和入侵行为识别的目的。主要内容包括以下几个方面：1.详细分析了被动入侵检测技术的国内外研究现状,并阐述了基于WIFI的被动入侵检测的研究意义。2.构建了室内实验仿真环境,采集了入侵目标在检测区域中站立和移动等行为下的RSSI数据,分析了不同行为下的RSSI变化情况,并对该数据进行预处理,形成样本特征集,根据提取的特征参数,采用SVM识别入侵行为。3.为了建立高效的检测模型,本文通过实验对比分析了不同核函数对识别效果的影响,并借助WEKA工具的特征选择功能对特征集进行了筛选。同时,利用网格搜索与交叉验证结合的方法对松弛变量和核函数参数进行了选取,验证了基于WIFI的被动入侵检测方法的可行性和有效性。实验结果表明该模型对入侵检测和判断入侵行为具有较好效果。
[Abstract]:Passive intrusion detection technology overcomes the problem that video and image detection technology is limited by environmental conditions, and detection targets do not need to carry wireless transceiver devices, monitoring in the elderly, With the development of WIFI technology, WiFi technology has many advantages such as high transmission speed, low cost and easy installation. The passive intrusion detection using WIFI can make use of the existing wireless network infrastructure more effectively, and can solve the traditional intrusion detection methods need special hardware support. The passive intrusion detection method based on WIFI is studied in this paper. The passive intrusion detection is mainly realized by wireless signals in the existing WIFI environment. By analyzing and processing received Signal Strength indicator (RSSI), the characteristics that can represent the intrusion behavior are extracted and classified. To achieve the purpose of passive intrusion detection and intrusion behavior recognition. The main contents include the following aspects: 1. The research status of passive intrusion detection technology at home and abroad is analyzed in detail. The research significance of passive intrusion detection based on WIFI is expounded. Secondly, the indoor simulation environment is constructed, and the RSSI data of the intrusion target under the behavior of standing and moving in the detection area are collected, and the RSSI changes under different behaviors are analyzed. The data is preprocessed to form the sample feature set. According to the extracted feature parameters, the intrusion behavior is identified by using SVM. 3. In order to establish an efficient detection model, the effect of different kernel functions on the recognition effect is compared and analyzed through experiments. The feature set is screened by the feature selection function of WEKA, and the relaxation variables and kernel function parameters are selected by combining grid search and cross-validation. The feasibility and effectiveness of the passive intrusion detection method based on WIFI are verified, and the experimental results show that the model has a good effect on intrusion detection and intrusion detection.
【学位授予单位】：合肥工业大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN92

【相似文献】