BIOS级身份认证系统的设计及实现

针对单机环境下通用PC终端常用身份认证方法存在安全隐患的缺点,提出了一种在计算机BIOS级进行用户身份认证的解决方案,用于增强主机系统的安全防护能力.在研究计算机BIOS系统安全功能增强技术的基础上,论述了基于USB Key的BIOS级身份认证系统的工作流程,详细分析了身份认证系统的组成结构及关键技术,并给出了相应的实现方法.实验结果表明,该身份认证系统能够较好地抵御典型的恶意攻击,满足PC终端安全防护的需要.     

基于ARM的IAP在线及远程升级技术

本文在阐述In-Application Programming(IAP)技术原理的基础上,以LPC2138为平台设计并实现了嵌入式微处理器ARM在线及远程升级方案。该方案方便快捷,摒弃了传统升级之前需要打开设备手工安装跳线等繁琐过程,并解决了升级过程意外中断、远程升级传输误码等关键技术问题。     

可穿戴设备数值型敏感数据本地差分隐私保护

针对数据服务器不可信时，直接收集可穿戴设备多维数值型敏感数据有可能存在泄露用户隐私信息的问题，通过引入本地差分隐私模型，提出了一种可穿戴设备数值型敏感数据的个性化隐私保护方案。首先，通过设置隐私预算的阈值区间，用户在区间内设置满足个人隐私需求的隐私预算，同时也满足了个性化本地差分隐私；其次，利用属性安全域将敏感数据进行归一化；最后，利用伯努利分布分组扰动多维数值型敏感数据，并利用属性安全域对扰动结果进行归一化还原。理论分析证明了该算法满足个性化本地差分隐私。实验结果表明该算法的最大相对误差（MRE）明显低于Harmony算法，在保护用户隐私的基础上有效地提高了不可信数据服务器从可穿戴设备收集数据的可用性。     

可穿戴设备拥有者转换协议的设计

近年来,可穿戴设备作为一类新型的智能设备正在改变着人们的生活方式。当这类设备处在多用户环境下时,应该考虑拥有者转换这一场景。为了解决可穿戴设备在拥有者转换场景下面临的安全隐私问题,设计了一个安全的可穿戴设备拥有者转换协议。该协议包括初始化、相互认证和密钥更新3个阶段。安全性分析表明本协议能够抵御传统的信道攻击和拥有者转换场景下的特殊攻击。实验分析表明所提方案具有较高的效率,能够适用于可穿戴设备环境。     

基于DSP／BIOS下TD-LTE协议栈系统框架的研究与实现

本文主要研究了DSP／BIOS实时嵌入式操作系统,深入探讨该系统的运行原理和特点,对其提供的关键系统内核对象进行详细的分析,并对系统中对象的配置和使用做了介绍。根据TD—LTE协议栈的系统构架和各子层的交互特点,结合工程实践,论述了基于DSP／BIOS系统下实现协议栈功能的方法,包括各子层的架构,层间通信,接口设计,同步机制等。尤其对实现过程中,各内核对象具体的参数配置进行了详细的分析,以便保证ID—LTE协议栈在该系统正常运行,对于更好的提高整个TD—LTE系统效率具有重要意义。     

基于ADS1293的穿戴式心电检测装置设计与实现

通过对穿戴式心电监护设备与低功耗技术的研究,设计了一款穿戴式心电检测装置。该装置将心电监护设备与背心服饰相结合,心电监护设备采用低功耗、高集成度的模拟前端芯片ADS1293与BLE蓝牙4.0低功耗无线模块进行心电信号的采集与传输,并结合低功耗MCU与电源管理模块可有效地降低系统功耗,然后通过Matlab GUI进行显示分析。通过实验测试验证,本装置体积小、功耗低、精度高,可在静息、行走和慢跑状态下长时间准确地采集心电信号,具有较高的可靠性、准确性和穿戴舒适性,可用于心脏疾病的远程监护与早期预警。     

The yoking-proof-based authentication protocol for cloud-assisted wearable devices

Along with the development of IoT applications, wearable devices are becoming popular for monitoring user data to provide intelligent service support. The wearable devices confront severe security issues compared with traditional short-range communications. Due to the limitations of computation capabilities and communication resources, it brings more challenges to design security solutions for the resource-constrained wearable devices in IoT applications. In this work, a yoking-proof-based authentication protocol (YPAP) is proposed for cloud-assisted wearable devices. In the YPAP, a physical unclonable function and lightweight cryptographic operators are jointly applied to realize mutual authentication between a smart phone and two wearable devices, and yoking-proofs are established for the cloud server to perform simultaneous verification. Meanwhile, Rubin logic-based security formal analysis is performed to prove that the YPAP has theoretical design correctness. It indicates that the proposed YPAP is flexible for lightweight wearable devices in IoT applications.     

可穿戴设备流数据的隐私保护发布

针对可穿戴设备流数据可能泄露个人隐私的问题,提出了一种基于自编码器和时频变换的隐私保护数据发布方法.通过分块离散余弦变换将滑动窗口数据变换为频谱数据,再通过自编码器实现脱敏变换,最后由重构的频谱数据逆变换回滑动窗口数据.利用预训练的活动识别与身份识别分类器评估自编码器输出结果的效用性和隐私性,通过多目标损失函数与反向传播更新自编码器权重.在Motion-Sense数据集上的实验结果表明,在重构数据上活动识别的F1-score由0.944降低至0.940,而身份识别的F1-score由0.908降低至0.673,重构加速度数据与原数据之间的均方误差为0.27.与同类算法相比,该算法能够更好地保留数据的效用性以及提高数据的安全性.     

Image recognition of sports training based on open IoT and embedded wearable devices

Computer Vision (CV) is a branch of artificial intelligence that is used to create a system for preparing a computer to understand and eliminate problem with artificial sense and images. Since video is a continuous image, or "shell" classification, this is more it can be used for recording. CV Visually they want to do something with the support framework visual inspection and learning model and high dynamic and visually complex simulation of human visibility part of the precision identification team use real-world images to transform real-world images. Supported vector machines can be used for incredible continuous order in order to be an asset, regular helper vector machine which has high stable performance. Therefore, depending on the installation of the adhesive hop field stabilizing the vector machine image, it is shown in this article. First, the first paragraph is used to prepare the screen confirmation model for deleting image elements based on population, input data acquisition technology invention, and two measurements each program, including carrier information, falls into the blur frame form of an image measurement carrier. At this point, can use the support vector machine 6 alone, the hop field of the support vector machine. The test results may confirm Support Vector Machine (SVM) capability in the hop field of insertion, show it in class 7, to realize an image that is clearer than conventional SVM.     

Athlete muscle measurement and exercise data monitoring based on embedded system and wearable devices

Wearable devices are considered to evaluate the performance of athletes during sports activities. Preliminary steps include recording motion variables at a sufficiently high sample rate throughout the test series. Load tracking sports science has become a very popular topic for athletes who are trained and competitive. Experiential research and process-based research have produced a rapid increase in the use of the dental approach by scientists and trainers to monitor the training load routinely and to capture the best way of capturing and interpreting data. This field has developed at such a rapid pace. In recent years, it has led to the development of new and novel model businesses that accurately place loads inside and outside to protect athletes and help them measure—protection from injury and poor health. Using Support Vector Machines (SVMs) to guide training and monitor athlete muscle measurement's well-being to detect any improvement in measures against poor team related to poor physical health effects all options and mechanisms required are proposed in the system analysis. Tracking application can be wearable technology that calculates the biomechanical and physiological parameters of athletes. Scored the second goal, that collaborative research opportunities between academic groups, sports medicine clinics, and sports team performance programs could help stream players across different sports fields use this technology. A supplementary biochemical profile and the use of wearable devices to track an athlete's mental intelligence were discussed. Key Indicator Value Not Available Using one wireless network connection can send other electronic devices at transmission speeds for reduction. Some experimental data have been reported on this method.