无线局域网指纹匹配身份认证仿真研究

摘要

对无线局域网指纹匹配身份进行准确的认证,可以保障无线局域网中身份认证过程中数据信息的安全性.应将对指纹图像纹理特征与系数特征构成指纹特征向量进行提取,获取存在平移和旋转关系的两幅指纹图像特征信息才能实现身份认证过程中的指纹匹配,但是传统方法通过小波去噪理论去除无线局域网中身份认证过程中语音信号中的环境噪声,结合马尔科夫模型进行预测信号特征匹配完成认证,但是不能准确提取纹图像纹理特征,存在认证误差大的问题.提出一种基于K均值理论的无线局域网指纹匹配身份认证方法.上述方法先利用K均值结合隶属度对身份认证中的指纹图像进行预处理,获取各类指纹图像的中心点,计算出指纹图像样本特征和各基准点特征的分块隶属度,依据最大隶属度原则获得指纹图像样本的分类结果,将分类的结果进行重构并提取指纹图像纹理特征与系数特征构成指纹特征向量,在此基础上对存在平移和旋转关系的两幅指纹图像特征信息利用限界盒方法进行全局匹配,实现身份认证过程中的指纹匹配,并完成指纹匹配身份认证.实验结果表明,所提方法进行无线局域网指纹匹配身份认证效果较好,且认证效率较优.%This research proposes an authentication method of fingerprint matching identity in WLAN based on K -means theory.The research preprocessed fingerprint images during identity authentication using the K-means integrated with membership degree and acquired central point of kinds of fingerprint image,then worked out block membership degree of sample feature of fingerprint image and each datum point feature,and obtained classification results of fingerprint image sample according to maximum membership principle.Moreover,the research reconstructed classification results and extracted feature vector consisting of texture feature and coefficient feature of fingerprint image.On that basis,the research carried out global matching for feature information of two fingerprint images which have the relations of translation and rotation using bounding box method.Experimental results show that the method has better effect during authentication of fingerprint matching identity and its authentication efficiency is optimum.