Graph in the Vault: Protecting Edge GNN Inference with Trusted Execution Environment
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866913701242601472 |
|---|---|
| author | Ding, Ruyi Xu, Tianhong Ding, Aidong Adam Fei, Yunsi |
| author_facet | Ding, Ruyi Xu, Tianhong Ding, Aidong Adam Fei, Yunsi |
| contents | Wide deployment of machine learning models on edge devices has rendered the model intellectual property (IP) and data privacy vulnerable. We propose GNNVault, the first secure Graph Neural Network (GNN) deployment strategy based on Trusted Execution Environment (TEE). GNNVault follows the design of 'partition-before-training' and includes a private GNN rectifier to complement with a public backbone model. This way, both critical GNN model parameters and the private graph used during inference are protected within secure TEE compartments. Real-world implementations with Intel SGX demonstrate that GNNVault safeguards GNN inference against state-of-the-art link stealing attacks with negligible accuracy degradation (<2%). |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2502_15012 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Graph in the Vault: Protecting Edge GNN Inference with Trusted Execution Environment Ding, Ruyi Xu, Tianhong Ding, Aidong Adam Fei, Yunsi Cryptography and Security Artificial Intelligence Wide deployment of machine learning models on edge devices has rendered the model intellectual property (IP) and data privacy vulnerable. We propose GNNVault, the first secure Graph Neural Network (GNN) deployment strategy based on Trusted Execution Environment (TEE). GNNVault follows the design of 'partition-before-training' and includes a private GNN rectifier to complement with a public backbone model. This way, both critical GNN model parameters and the private graph used during inference are protected within secure TEE compartments. Real-world implementations with Intel SGX demonstrate that GNNVault safeguards GNN inference against state-of-the-art link stealing attacks with negligible accuracy degradation (<2%). |
| title | Graph in the Vault: Protecting Edge GNN Inference with Trusted Execution Environment |
| topic | Cryptography and Security Artificial Intelligence |
| url | https://arxiv.org/abs/2502.15012 |