ARCSim

ARCSim数据集是由德黑兰大学电气与计算机工程系创建，用于分析和预测电子设备中的指令流。该数据集包含163条指令，覆盖了AVR微控制器的多种指令类型，每条指令包含400至200个样本点。数据集的创建过程涉及随机指令生成、编译、运行时标签提取等步骤，旨在通过深度神经网络模型提高逆向工程的准确性和效率。ARCSim数据集主要应用于硬件安全领域，特别是侧信道攻击和逆向工程，以解决低分辨率数据下的攻击准确性问题。

The ARCSim dataset was created by the Department of Electrical and Computer Engineering, University of Tehran, for analyzing and predicting instruction flows in electronic devices. It comprises 163 instructions covering multiple instruction types of AVR microcontrollers, with 400 to 200 sample points per instruction. The dataset creation process involves steps such as random instruction generation, compilation, and runtime label extraction, aiming to enhance the accuracy and efficiency of reverse engineering via deep neural network models. The ARCSim dataset is primarily applied in the field of hardware security, particularly side-channel attacks and reverse engineering, to address the challenge of attack accuracy under low-resolution data.