Oblivious RAM (ORAM) is a security primitive to prevent the memory access pattern leakage. By adding redundant memory accesses, ORAM prevents attackers from revealing the patterns in the access sequences by breaking the locality. Meanwhile, security protection comes with the cost of performance loss. Therefore, from the architectural perspective, ORAM is still too expensive to be adopted on scalable computing systems.
In this talk, I will talk about our recent works which enable scalable and practical ORAM protection with architectural enhancement. I will first introduce CP-ORAM, which includes a flexible and effective bandwidth allocation algorithm that optimizes the resource sharing between secure and non-secure applications. Then I will introduce D-ORAM, which reduces the co-run interference by leveraging the buffer-on-board memory architecture to delegate the ORAM primitives. Lastly, I will introduce H-ORAM, a hybrid ORAM interface that addresses massive performance degradation when overflowing the user data to storage.
Rujia Wang is an Assistant Professor in the Computer Science Department at Illinois Institute of Technology. She received her Ph.D. degree from the University of Pittsburgh in 2018 and B.E. from Zhejiang University in 2013. Her research experience spans across multiple areas in computer architecture and systems, with a focus on the design and optimization of new memory hierarchy and systems for better performance, scalability, security, and reliability. Her work has been published in top conferences in computer architecture area.
downloadFile