Privacy-Preserving Approximate Nearest Neighbor Search on High-Dimensional Data

TL;DR

This paper presents a novel privacy-preserving approximate nearest neighbor search method on high-dimensional data that ensures data privacy, improves efficiency, and maintains accuracy, primarily executing on a single cloud server.

Contribution

The paper introduces a new encryption technique called distance comparison encryption and a privacy-preserving index combining state-of-the-art methods with approximate distance computation.

Findings

Accelerates PP-ANNS by up to 1000 times compared to existing methods.

Maintains search accuracy while enhancing privacy and efficiency.

Demonstrates superior performance through extensive experiments.

Abstract

In the era of cloud computing and AI, data owners outsource ubiquitous vectors to the cloud, which furnish approximate $k$-nearest neighbors ($k$-ANNS) services to users. To protect data privacy against the untrusted server, privacy-preserving $k$-ANNS (PP-ANNS) on vectors has been a fundamental and urgent problem. However, existing PP-ANNS solutions fall short of meeting the requirements of data privacy, efficiency, accuracy, and minimal user involvement concurrently. To tackle this challenge, we introduce a novel solution that primarily executes PP-ANNS on a single cloud server to avoid the heavy communication overhead between the cloud and the user. To ensure data privacy, we introduce a novel encryption method named distance comparison encryption, facilitating secure, efficient, and exact distance comparisons. To optimize the trade-off between data privacy and search performance, we design a privacy-preserving index that combines the state-of-the-art $k$-ANNS method with an approximate distance computation method. Then, we devise a search method using a filter-and-refine strategy based on the index. Moreover, we provide the security analysis of our solution and conduct extensive experiments to demonstrate its superiority over existing solutions. Based on our experimental results, our method accelerates PP-ANNS by up to 3 orders of magnitude compared to state-of-the-art methods, while not compromising the accuracy.