基于混沌测量矩阵的生物哈希密文语音检索  

Biohashing encrypted speech retrieval based on chaotic measurement matrix

作　　者：黄羿博 王勇 张秋余[2] 陈腾飞 HUANG Yibo;WANG Yong;ZHANG Qiuyu;CHEN Tengfei(College of Physics and Electronic Engineering,Northwest Normal University,Lanzhou 730070,China;School of Computer and Communication,Lanzhou University of Technology,Lanzhou 730050,China)

机构地区：[1]西北师范大学物理与电子工程学院,甘肃兰州730070 [2]兰州理工大学计算机与通信学院,甘肃兰州730050

出　　处：《华中科技大学学报（自然科学版）》2020年第12期32-37,共6页Journal of Huazhong University of Science and Technology(Natural Science Edition)

基　　金：国家自然科学基金资助项目(61862041);甘肃省青年科学技术基金资助项目(1606RJYA274)。

摘　　要：为了解决现存语音检索系统中明文数据的泄露问题,提高语音检索性能、生物特征模板的安全性和隐私性,提出了一种基于混沌测量矩阵的生物哈希密文语音检索算法.首先,用户端对语音进行分类,再分发与类为单一映射的密钥,通过密钥生成358位的Rossler混沌测量矩阵,并使用该矩阵对语音特征进行特征变换,进一步二值化生成语音的哈希索引;然后,通过改进的sha256算法对语音文件进行加密;最后,将哈希索引和加密语音送至云端.实验结果表明:提出的算法不仅能防止明文泄露,而且具有良好的鲁棒性、区分性和检索性能;与此同时,生物特征模板具备良好的多样性、可撤销性、安全性和隐私性.In order to solve the problem of plaintext data leakage in the existing speech retrieval system,and improve the performance of speech retrieval,the security and privacy of biometric template,a biohashing encrypted speech retrieval algorithm was proposed based on Rossler chaotic measurement matrix.First,the speech was classified by the client,and redistribute key with class as single mapping.A 358 bit Rossler chaotic measurement matrix was generated by the key,the matrix was used to transform the speech features,and the hash index of speech was generated by binarization.Then the speech file was encrypted using the improved sha256 algorithm.Finally,hash index and encrypted speech were sent to the cloud.The experimental results show that this algorithm not only can effectively prevent plaintext leakage,but also has good robustness,discrimination and retrieval performance.At the same time,biometric template has good diversity,revocability,security and privacy.

关 键 词：密文语音检索 生物特征模板 生物哈希 Rossler混沌测量矩阵 改进的sha256算法 

分 类 号：TP391.3[自动化与计算机技术—计算机应用技术] TN912.3[自动化与计算机技术—计算机科学与技术]