近年来，侧信道分析是对物联网系统的密码模块、可信执行环境TEE威胁最大的攻击之一。现有物联网设备资源受限，其抗侧信道防护能力通常十分薄弱。本项目基于申请团队四名骨干分别主持的已结题自然基金中对ARM、嵌入式、IC卡等设备的侧信道分析研究基础，进一步探索面向物联网设备的远距离Cache计时分析技术、中距离声音和电磁分析技术，并将其应用于物联网产品的安全性测评。.本项目拟研究TEE隔离技术、物联网设备侧信道攻击的一般化机理，针对应用TEE技术的移动智能终端与同驻虚拟机环境下的Cache计时分析、声音与电磁信号的采集/处理/分析等技术开展研究，突破Cache/声音/电磁泄露等侧信道信息远程采集、跨虚拟机Cache计时信息探测、基于机器学习的侧信道分析等关键问题，结合人工智能算法改进侧信道信息分析算法，并通过自主实现物联网Cache计时分析、声音分析、电磁分析实验平台，对所提新算法进行实验验证。

In recent years, side-channel analyses show significant threat on cryptographic module and trusted execution environment of IoT systems. Due to the limited resources, defense of IoT devices is very weak. In this project, based on the previous side-channel analysis researches on ARM, embedded systems, and IC cards, we focus on the remote cache timing analysis, acoustic and electromagnetic analysis on IoT devices, and apply these technologies to the security evaluation of IoT Devices..We conclude the common principle of TEE isolation and Side-channel Attacks on IoT devices, and study Cache timing analysis on smart mobile device with TEE and virtual machine environment, and acquisition, process, and analysis of Cache timing, sound and electromagnetic signal. The open problems such as remote acquisition of Cache, acoustic and electromagnetic signal, side-channel information probe of Cache among the virtual machine, and side-channel analysis based on machine learning are broken through. Some artificial intelligence algorithms are employed to improve the side-channel analysis algorithm. We build the platforms of Cache timing, acoustic, and EM analysis, and verify the proposed algorithms.