A GPU-Accelerated Framework for Standard White-Box Cryptographic Algorithms in Unattended IoT Devices

White-box cryptography is widely used in Internet of Things (IoT) devices to ensure data confidentiality. However, the traditional implementations of white-box cryptographic algorithms (WBCAs) are inefficient and impractical for IoT devices that require fast encryption and decryption of large amounts of data. To address this challenge, we propose a framework that can accelerate WBCAs employed by IoT devices with Graphics Processing Units (GPUs). Our framework leverages the inherent multithreading capabilities of GPUs to simultaneously perform a large number of table lookups required in WBCAs. Additionally, we employ pipelined execution strategies to enhance performance further. To demonstrate the effectiveness of our framework, we apply it to two well-known WBCAs and evaluate its performance on two IoT devices. The experimental results show that our framework can improve performance by up to 70 times compared with the execution on a single-threaded CPU.