基于对抗神经网络的医院科研数据混合加密系统设计  

Design of Hospital Scientific Research Data Hybrid Encryption System Based on Adversarial Neural Network

作　　者：王玉珏 WANG Yu-jue(The Fourth Affiliated Hospital of Harbin Medical University,Harbin 150026 China)

机构地区：[1]哈尔滨医科大学附属第四医院,黑龙江哈尔滨150026

出　　处：《自动化技术与应用》2023年第10期85-87,120,共4页Techniques of Automation and Applications

摘　　要：针对医疗科研数据加密系统解密误差大等问题,设计一种混合加密系统。该系统实现数据资源的独立分配计算,优化混合加密算法。设置不同长度的密钥对设计的系统和基于深度学习的加密系统进行测试,实验结果表明,当密钥长度较长时,设计系统解密平均错误bit数量明显低于传统系统,验证了设计系统在实际应用中的可靠性。In this paper,a hybrid encryption system is designed for the decryption error of medical research data encryption system.This system realizes the independent allocation calculation of data resources and optimizes the hybrid encryption algorithm.During the test,different lengths are keys to test the designed system and the encryption system based on deep learning.The experimental results show that when the key length is long,the average decryption error bit of the design system is significantly lower than that of the traditional system,so as to verify the reliability of the design system in practical application.

关 键 词：对抗神经网络 混合加密算法 数据脱敏 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程] TP309.7[自动化与计算机技术—控制科学与工程]