摘要
针对边缘环境下的图像隐私泄露和计算效率问题,提出一种边缘协作的轻量级安全区域建议网络(SecRPN)。基于加性秘密共享方案设计一系列安全计算协议,由2台非共谋边缘服务器协作执行安全特征处理、安全锚变换、安全边界框修正、安全非极大值抑制等计算模块。理论分析证明了SecRPN的正确性和安全性,实际性能评估表明,计算和通信开销均远优于现有工作。
Aiming at the problem of image privacy leakage and computing efficiency in edge environment,a lightweight and secure region proposal network(SecRPN)was proposed.A series of secure computing protocols were designed based on the additive secret sharing scheme.Two non-collusive edge servers cooperate to perform calculation modules such as secure feature processing,secure anchor transformation,secure bounding-box correction,and secure non-maximum suppression.Theoretical analysis guarantees the correctness and security of SecRPN.The actual performance evaluation shows that Sec RPN is outstanding in the computational cost and communication overhead compared with the existing works.
作者
熊金波
毕仁万
陈前昕
刘西蒙
XIONG Jinbo;BI Renwan;CHEN Qianxin;LIU Ximeng(College of Mathematics and Informatics,Fujian Normal University,Fuzhou 350117,China;Fujian Provincial Key Laboratory of Network Security and Cryptology,Fujian Normal University,Fuzhou 350007,China;College of Mathematics and Computer Science,Fuzhou University,Fuzhou 350108,China;Fujian Provincial Key Laboratory of Network System Information Security,Fuzhou University,Fuzhou 350108,China)
出处
《通信学报》
EI
CSCD
北大核心
2020年第10期188-201,共14页
Journal on Communications
基金
国家自然科学基金资助项目(No.61872088,No.U1905211,No.U1804263,No.61702105,No.61872090)
福建省自然科学基金资助项目(No.2019J01276)
贵州省公共大数据重点实验室开放课题基金资助项目(No.2019BDKFJJ004)
陕西省网络与系统安全重点实验室开放课题基金资助项目(No.NSSOF1900104)
广东省数据安全与隐私保护重点实验室开放课题基金资助项目(No.2017B03031004-12)。
关键词
边缘协作
区域建议网络
目标检测
加性秘密共享
安全计算协议
edge-collaborative
region proposal network
object detection
additive secret sharing
secure computing protocol
