R&D on privacy-preserving outsourced computing in cloud computing
| Project/Area Number |
15K00028
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Multi-year Fund |
|---|
| Section | 一般 |
|---|
| Research Field |
Theory of informatics
|
|---|
| Research Institution | Kobe University (2018)
National Institute of Information and Communications Technology (2015-2017) |
|---|
Principal Investigator |
Wang Lihua 神戸大学, 工学研究科, 特命准教授 (00447228)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
林 卓也 国立研究開発法人情報通信研究機構, サイバーセキュリティ研究所セキュリティ基盤研究室, 研究員 (70739995)
|
|---|
| Research Collaborator |
MAMBO Masahiro
WANG Licheng
AONO Yoshinori
LE Trieu Phong
|
|---|
| Project Period (FY) |
2015-04-01 – 2019-03-31
|
| Project Status |
Completed (Fiscal Year 2018)
|
| Budget Amount *help |
¥4,680,000 (Direct Cost: ¥3,600,000、Indirect Cost: ¥1,080,000)
Fiscal Year 2018: ¥1,040,000 (Direct Cost: ¥800,000、Indirect Cost: ¥240,000)
Fiscal Year 2017: ¥910,000 (Direct Cost: ¥700,000、Indirect Cost: ¥210,000)
Fiscal Year 2016: ¥1,300,000 (Direct Cost: ¥1,000,000、Indirect Cost: ¥300,000)
Fiscal Year 2015: ¥1,430,000 (Direct Cost: ¥1,100,000、Indirect Cost: ¥330,000)
|
|---|
| Keywords | クラウドセキュリティ / 代理再暗号 / 準同型暗号 / 秘密計算 / プライバシー保護データマイニング / 暗号 / 代理計算 / セキュア計算 / 代理人再暗号 / 結託攻撃 / 安全性帰着方法 |
|---|
| Outline of Final Research Achievements |
In this research, we studied the privacy-preserving outsourced computation in cloud computing and its applications in machine learning. In particular, we proposed several protocols for the inner product, matrix product, and integer comparison on ciphertexts. All the proposed protocols are secure against attacks by quantum computers, and they are also efficient according to the extensive experimental results. Based on these protocols, we also proposed several privacy-preserving outsourced machine learning schemes, including Naive Bayes classification and 3-layer neural networks. The detailed security analyses showed that these schemes would not reveal any information to the cloud or others during the process, and the experimental results also demonstrated that they are relatively efficient. With the above research results, we have published seven papers in international conferences and journals, and have applied for a patent.
|
| Academic Significance and Societal Importance of the Research Achievements |
クラウドコンピューティングに代表される、計算処理をサーバに委任する「代理計算」においては、計算対象データのサーバへの開示が必要なため、データ漏えい等の対策が必須である。本研究ではこの課題の解決策として、暗号技術を応用した安全な代理計算技術の研究開発を行っており、データを暗号化したまま計算を行うことができる「準同型暗号」において、ベクトル内積などの計算プリミティブの効率的な方式の開発や、応用例としてニューラルネットワーク推論処理などの機械学習アルゴリズムの安全な代理計算を構築した。これらの結果は、特定の処理に特化した代理計算については、データを安全に保ちつつ効率良く計算ができることを示している。
|
Report
(5 results)
Research Products
(24 results)
-
-
-
-
-
-
-
-
-
-
-
[Journal Article] A Generic yet Efficient Method for Secure Inner Product2017
Author(s)
Lihua Wang, Takuya Hayashi, Yoshinori Aono, Le Trieu Phong
-
Journal Title
Proceedings of NSS2017- 11th International Conference on Network and System Security, Springer International Publishing
Volume: LNCS 10394
Pages: 217-232
DOI
ISBN
9783319647005, 9783319647012
Related Report
Peer Reviewed / Open Access
-
-
-
-
-
-
-
-
-
-
-
-
-
