Towards Practical Oblivious RAM

TL;DR

This paper advances practical Oblivious RAM by introducing a new construction that significantly reduces overhead and employs a novel partitioning and eviction technique for improved efficiency.

Contribution

It presents a practical O-RAM construction with substantially lower overhead and introduces a novel partitioning and background eviction method for better performance.

Findings

Achieves 20X-35X amortized overhead for 1TB O-RAM

Approximately 63 times faster than previous schemes

Introduces a novel partitioning and eviction technique

Abstract

We take an important step forward in making Oblivious RAM (O-RAM) practical. We propose an O-RAM construction achieving an amortized overhead of 20X-35X (for an O-RAM roughly 1 terabyte in size), about 63 times faster than the best existing scheme. On the theoretic front, we propose a fundamentally novel technique for constructing Oblivious RAMs: specifically, we partition a bigger O-RAM into smaller O-RAMs, and employ a background eviction technique to obliviously evict blocks from the client-side cache into a randomly assigned server-side partition. This novel technique is the key to achieving the gains in practical performance.