基于声音识别的计算机网络通信数据加密方法  

Data Encryption Method for Computer Network Communication Based on Voice Recognition

作　　者：翁梦倩 WENG Mengqian(Jiangxi Technical College of Manufacturing,Nanchang 330095,China)

机构地区：[1]江西制造职业技术学院,江西南昌330095

出　　处：《电声技术》2024年第11期51-53,共3页Audio Engineering

摘　　要：为提升计算机网络通信数据的安全性,提出一种基于声音识别技术的加密方法。该方法通过提取语音信号的梅尔频率倒谱系数(Mel-Frequency Cepstrum Coefficient,MFCC),结合混沌映射与256位安全散列算法(256-bits Secure Hash Algorithm,SHA-256)生成加密密钥,并使用Salsa20流密码算法加密数据。实验结果显示,该方法不仅提高了密钥的安全性,而且在加密速度上优于传统256位高级加密标准(256-bits Advanced Encryption Standard,AES-256)和Blowfish算法,密文熵接近理想值,通过了美国国家标准与技术研究院(National Institute of Standards and Technology,NIST)随机性测试,适用于实时网络通信数据保护。In order to improve the security of computer network communication data,an encryption method based on voice recognition technology is proposed.In this method,the Mel-Frequency Cepstrum Coefficient(MFCC)of speech signal is extracted,and the encryption key is generated by combining chaotic mapping with 256-bits Secure Hash Algorithm(SHA-256),and the data is encrypted by Salsa20 stream cipher algorithm.The experimental results show that this method not only improves the security of the key,but also is superior to the traditional 256-bits Advanced Encryption Standard(AES-256)and Blowfish algorithm in encryption speed,and the ciphertext entropy is close to the ideal value,which has passed the National Institute of Standards and Technology(NIST)randomness test and is suitable for real-time network communication data protection.

关 键 词：声音识别 数据加密 Salsa20 

分 类 号：TP393.08[自动化与计算机技术—计算机应用技术]