Abstract

Properly stored, shared, and utilized, data can create immense value—a fact increasingly recognized by individuals, companies, and governments alike. This potential is not only harnessed by those with benign intentions but also exploited by cybercriminals. The recent surge in cyber-attacks has impacted millions of users, resulting in losses amounting to billions of dollars. This trend underscores the urgent need for the development and proactive deployment of robust data security measures. To combat persistent and pervasive security threats, the encryption of data at all times is emerging as a critical strategy and, in some cases, a regulatory imperative. However, traditional data encryption methods often preclude meaningful data operations.

Fig. 1: Our proposed architecture for encrypted database

In this research proposal, we outline plans to construct a fully functional encrypted database leveraging secure hardware enclaves. We are confident that such encrypted databases will find application in numerous real-life data services, provided our design is realized through preliminary evaluation procedures. Despite their promise, current approaches to searching encrypted data often fall short in terms of security, efficiency, or query expressiveness. Our research will explore a new hardware-assisted security paradigm anticipated to achieve a harmonious balance of all critical properties. This paradigm aims to address the excessive information leakage inherent in cryptographic techniques such as searchable encryption and property-preserving encryption, while also enhancing practical performance and functionality.

Publication

• Yi Liu, Cong Wang, and Xingliang Yuan, "BadSampler: Harnessing the Power of Catastrophic Forgetting to Poison Byzantine-robust Federated Learning", In ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (ACM KDD'24), August 25-29, 2024.


• Leqian Zheng, Lei Xu, Cong Wang, Sheng Wang, Yuke Hu, Zhan Qin, Feifei Li, Kui Ren, "SWAT: A System-Wide Approach to Tunable Leakage Mitigation in Encrypted Data Stores", In International Conference on Very Large Databases (VLDB'24), August 26-30, 2024.


• Xiang Zheng, Xingjun Ma, Shen Chao, and Cong Wang, "Constrained Intrinsic Motivation for Reinforcement Learning", In International Joint Conference on Artificial Intelligence (IJCAI'24), August 3-9, 2024.

People