PrivGNN: High-Performance Secure Inference for Cryptographic Graph Neural Networks

TL;DR

PrivGNN introduces a lightweight cryptographic scheme that significantly accelerates secure inference for graph neural networks in cloud environments, ensuring data privacy without sacrificing accuracy.

Contribution

The paper presents novel 2PC protocols and a hybrid cryptographic scheme that substantially improve the efficiency of secure GNN inference compared to existing solutions.

Findings

Achieves 1.5x to 1.7x speedup for linear layers

Achieves 2x to 15x speedup for non-linear layers

Demonstrates 1.3x to 4.7x faster secure predictions across datasets

Abstract

Graph neural networks (GNNs) are powerful tools for analyzing and learning from graph-structured (GS) data, facilitating a wide range of services. Deploying such services in privacy-critical cloud environments necessitates the development of secure inference (SI) protocols that safeguard sensitive GS data. However, existing SI solutions largely focus on convolutional models for image and text data, leaving the challenge of securing GNNs and GS data relatively underexplored. In this work, we design, implement, and evaluate $\sysname$, a lightweight cryptographic scheme for graph-centric inference in the cloud. By hybridizing additive and function secret sharings within secure two-party computation (2PC), $\sysname$ is carefully designed based on a series of novel 2PC interactive protocols that achieve $1.5\times \sim 1.7\times$ speedups for linear layers and $2\times \sim 15\times$ for non-linear layers over state-of-the-art (SotA) solutions. A thorough theoretical analysis is provided to prove $\sysname$'s correctness, security, and lightweight nature. Extensive experiments across four datasets demonstrate $\sysname$'s superior efficiency with $1.3\times \sim 4.7\times$ faster secure predictions while maintaining accuracy comparable to plaintext graph property inference.