基于大规模矩阵Jordan分解的外包计算  

Outsourcing computing of large matrix Jordan decomposition

作　　者：吴宏锋[1] 闫晶晶 Wu Hongfeng;Yan Jingjing(College of Science,North China University of Technology,Beijing 100144)

机构地区：[1]北方工业大学理学院

出　　处：《网络空间安全》2019年第2期89-95,共7页Cyberspace Security

基　　金：国家自然科学基金(项目编号:61370187);北京市教委科技计划项目(项目编号:KM201510009013)

摘　　要：目前,外包计算已成为减轻用户庞大计算量的重要策略之一。针对大规模矩阵的Jordan分解需要用户付出大量的计算资源问题。文章设计了一个安全、结果可验证、高效的外包协议,达到了节省用户计算资源的目的。通过线性变换、元素的重排列对原始矩阵进行加密,保护了用户隐私信息,并运用高效的验证算法对云端返回的结果进行了高效验证。文章通过计算复杂度分析,验证了该协议的高效性。At present,outsourcing computing has become an important way to reduce the large amount of computing by users.The Jordan decomposition of large-scale matrix requires users to pay a lot of computing resources.In order to save user's computing resources,the article designs a secure,verifiable and efficient outsourcing protocol for Jordan decomposition of large-scale matrix.The original matrix is encrypted by linear transformation and element rearrangement to protect the user's privacy information.Efficient verification algorithm is used to verify the results returned from the cloud.Finally,the computational complexity is analyzed and the efficiency of the protocol is verified.

关 键 词：外包计算 JORDAN分解 可验证性 

分 类 号：TP309[自动化与计算机技术—计算机系统结构]