Core Concepts
RAGE introduces a novel, lightweight CFA approach for embedded devices, addressing limitations of existing schemes.
Abstract
Introduction: Discusses the limitations of current Control-Flow Attestation (CFA) schemes.
RAGE Approach: Introduces RAGE as a solution for embedded devices, utilizing Graph Neural Networks.
Evaluation: Demonstrates RAGE's effectiveness in detecting real-world and synthetic attacks on various software benchmarks.
System Model: Describes the system model and threat scenario for attestation.
Data Collection: Details the methodology used to collect execution traces on different platforms.
Datasets: Includes datasets from Embench benchmarks, OpenSSL library, real-world attacks, and synthetic trace generation algorithms.
Model Training: Outlines the training process of the VGAE model for attestation.
Data Preprocessing: Explains the feature extraction phase to prepare raw data for machine learning steps.
Stats
RAGEは、埋め込みデバイス向けの新しい軽量CFAアプローチを導入します。
現在のCFAスキームの制限に対処するためにRAGEが開発されました。