基于稀疏训练点和指纹重构的室内可见光三维定位算法

目前构建基于机器学习的室内可见光定位模型主要依赖于光电二极管和指纹数量,为了降低指纹采集的复杂度,提高定位精度,提出一种基于指纹矩阵稀疏重构的室内三维可见光定位算法。该算法利用极限学习机训练稀疏采样点,采用奇异值分解和交替方向乘子法求解稀疏指纹矩阵的重构问题。该算法可以有效降低指纹的采样率,同时可以基于极限学习机算法较强的泛化能力提高定位速度和定位精度。在此基础上,由于可见光的多径反射等因素的影响,定位区域的边界定位误差大于内部定位误差,通过引入一种边界修正定位算法,可以有效降低边界定位误差。仿真和实验结果表明,与传统的机器学习算法相比,该算法在减少其所需指纹数量的同时,具有更高的定位速度和精度。     

基于改进粒子滤波的RFID室内目标定位算法

针对于LANDMARC算法的RFID室内定位精度受传输路径影响严重,直接采用粒子滤波自适应性差的问题,提出一种基于改进粒子滤波的RFID室内定位算法。该算法首先利用极限学习机(ELM)拟合阅读器接收信号强度与标签距离之间的非线性关系,构建信号传输模型,筛选邻近标签集;然后采用自适应学习因子优化粒子滤波过程,提高粒子全局寻优能力和收敛速度。仿真实验结果表明,该算法能够有效实现待测标签的RFID室内定位,且定位精度较高,收敛速度较快。     

基于机器学习的可见光通信信号处理研究现状

随着无线通信领域的发展,具有诸多优点的可见光通信(VLC)已经发展成为了一种具有广阔前景的通信手段。然而,可见光通信中的各种非线性效应会给其信号处理带来诸多的困难,并恶化系统的性能。机器学习在解决非线性问题方面具有很大的优势和潜力,结合机器学习算法的可见光通信技术必然具有巨大的研究价值。已有研究表明,传统的机器学习算法如K-means、DBSCAN以及支持向量机(SVM)等在预均衡、后均衡、抗系统抖动,以及相位纠正等方面均有很好的表现。而深度神经网络(DNN)则因为其强大的非线性拟合能力能够更进一步提升VLC系统的性能。对以上几种方法进行了分析和介绍,并对其在可见光通信信号处理领域的应用进行了分析与总结,希望可以为机器学习解决可见光通信方面的各种非线性问题提供参考。     

Recent advances in deep learning‐based side‐channel analysis

As side‐channel analysis and machine learning algorithms share the same objective of classifying data, numerous studies have been proposed for adapting machine learning to side‐channel analysis. However, a drawback of machine learning algorithms is that their performance depends on human engineering. Therefore, recent studies in the field focus on exploiting deep learning algorithms, which can extract features automatically from data. In this study, we survey recent advances in deep learning‐based side‐channel analysis. In particular, we outline how deep learning is applied to side‐channel analysis, based on deep learning architectures and application methods. Furthermore, we describe its properties when using different architectures and application methods. Finally, we discuss our perspective on future research directions in this field.     

可见光室内定位技术研究进展

可见光赋能的定位技术具有无电磁辐射、成本低、精度高、安全性高、不易受电磁干扰等优势,故近年来可见光室内定位技术(Visible Light Indoor Positioning, VLIP)受到了学界和业界的广泛关注。为了展示VLIP的研究进展和发展全貌,阐述了VLIP的系统架构及相关通信技术,对VLIP基础定位方法及其相关辅助定位技术进行了详细地梳理、分类和对比分析,针对VLIP当前面临的挑战分析了潜在的解决方案,展望了VLIP未来研究方向。     

面向机器学习的隐私保护关键技术研究综述

随着信息通信技术的发展,机器学习已经成为多个研究领域与垂直行业必不可少的技术工具。然而,机器学习所需数据中往往包含了大量的个人信息,使其隐私保护面临风险与挑战,受到了越来越多的关注。对现有机器学习下隐私保护法规政策与标准化现状进行梳理,对适用于机器学习的隐私保护技术进行详细介绍与分析。隐私保护算法通常会对数据质量、通信开支与模型表现等造成影响,因此对于隐私保护算法的评估应当进行多维度的综合评估。总结了适用于机器学习应用的隐私保护性能评估指标,并指出隐私保护需要考虑对数据质量、通信开支以及模型准确率等之间的影响。     

基于指纹算法的无线室内定位技术

随着人们对基于位置的服务（LBS）需求日益增大,室内定位逐渐成为用户定位领域的研究热点,而指纹定位因具有定位精度高、普适性强和无需额外设备等优点而受到大多数研究者的青睐。首先详细综述了各种主流室内位置指纹定位技术的定位原理,然后归纳了现有的室内定位算法的原理及发展现状,最后搭建楼宇通道的测试场景,对各种典型室内定位算法进行测试验证和比较分析,为室内定位技术的研究与应用人员提供参考。     

基于深度学习的信道估计技术研究进展

信道估计是接收机基带信号处理的关键,直接决定了无线通信系统的通信服务质量。传统的信道估计方法已经不能满足日益复杂和个性化的现代通信需求,同时人工智能技术特别是深度学习已被应用于无线通信物理层并带来了良好的通信性能增益。为系统地总结上述研究成果,并探讨未来的技术发展趋势,从数据驱动和模型驱动两方面分别对基于深度学习的信道估计方法进行了分析和归纳,并且描述了其中代表性算法,最后探讨了基于深度学习的信道估计的研究挑战与趋势。     

ILONA: indoor localization and navigation system

Indoor positioning is a hot topic these days and there is a growing need for it in public buildings such as airports, hospitals, universities or shopping malls. Indoor positioning systems should be accurate, easily available for the users, with low installation and maintenance cost, which makes development challenging. Existing systems are based on various technologies such as ultrasonic, RFID, WiFi or light encoding. Moreover, these systems are tailored to a given environment and usually rely on a single technology. This paper presents the indoor localization and navigation (ILONA) System, a flexible hybrid indoor positioning and navigation framework. The ILONA System was not designed to be a solution for a single indoor positioning task but to be a standard core component of various systems. The ILONA System provides easily available positioning and navigation services for the end users. The system can manage data from the most commonly available sensors of modern smart phones. Thus, the ILONA System can perform positioning based on various technologies. ILONA System can be established at low cost because it only requires a connection between the server and the clients and WLAN is usually available. Hence, the presented ILONA System provides a widely available, hybrid indoor positioning framework at low cost to the developers of other indoor positioning solutions.     

基于机器学习的图像分割算法研究

图像分割是计算机视觉研究中重要的一部分,其主要目的是在图像中将兴趣域目标与背景分割,关系到后续的目标识别、图像理解等操作的准确性.经过几十年的发展,许多优秀的图像分割的方法被提出.机器学习是当今时代的研究热点,基于深度卷积神经网络等机器学习的图像分割研究进展迅速.总结介绍了应用于图像分割的几种典型机器学习方法,分析比较了相关的分割原理步骤、优缺点和发展现状.最后分析了基于机器学习的图像分割算法的发展方向.     

网络欺凌检测综述

网络欺凌在社交媒体平台的日益泛滥引起了研究者的广泛关注,社会科学和计算机科学研究者从不同的角度对该问题进行了研究与探讨.为梳理这些研究,本论文对社会科学领域和计算机领域在网络欺凌方面的研究进行了调查分析.首先概述了网络欺凌的基本研究内容和网络欺凌特征,重点讨论了各种用于网络欺凌检测的机器学习方法,包括基于监督学习、基于弱监督学习、基于预设规则和深度学习算法,随后总结了12个现有的网络欺凌检测数据集和常用的检测性能评价指标,最后对基于异构信息网络、融合多种辅助信息和结合心理学特征的欺凌检测方法等进行了展望.     

A new method for improving Wi-Fi-based indoor positioning accuracy

Wi-Fi- and smartphone-based positioning technologies are playing a more and more important role in location-based service industries due to the rapid development of the smartphone market. However, the low positioning accuracy of these technologies is still an issue for indoor positioning. To address this problem, a new method for improving the indoor positioning accuracy was developed. The new method initially used the nearest neighbour (NN) algorithm of the fingerprinting method to identify the initial position estimate of the smartphone user. Then two distance correction values in two roughly perpendicular directions were calculated by the path loss model based on the two signal strength indicator values observed. The systematic error from the path loss model were eliminated by differencing two model-derived distances from the same access point. The new method was tested and the results compared and assessed against that of the commercial Ekahau RTLS system and the NN algorithm. The preliminary results showed that the positioning accuracy has been improved consistently after the new method was applied and the root mean square accuracy improved to 3.3 m from 3.8 m compared with the NN algorithm.     

Location determination using WiFi fingerprinting versus WiFi trilateration

Many applications in the area of location-based services and personal navigation require nowadays the location determination of a user not only in an outdoor environment but also an indoor. Typical applications of location-based services (LBS) mainly in outdoor environments are fleet management, travel aids, location identification, emergency services and vehicle navigation. LBS applications can be further extended if reliable and reasonably accurate three-dimensional positional information of a mobile device can be determined seamlessly in both indoor and outdoor environments. Current geolocation methods for LBS may be classified as GNSS-based, cellular network-based or their combinations. GNSS-based methods rely very much on the satellite visibility and the receiver-satellite geometry. This can be very problematic in dense high-rise urban environments and when transferring to an indoor environment. Especially, in cities with many high-rise buildings, the urban canyon will greatly affect the reception of the GNSS signals. Moreover, positioning in the indoor/outdoor transition areas would experience signal quality and signal reception problems, if GNSS systems alone are employed. The authors have proposed the integration of GNSS with wireless positioning techniques such as WiFi and UWB. In the case of WiFi positioning, the so-called fingerprinting method based on WiFi signal strength observations is usually employed. In this article, the underlying technology is briefly reviewed, followed by an investigation of two WiFi-positioning systems. Testing of the system is performed in two localisation test beds, one at the Vienna University of Technology and another one at the Hong Kong Polytechnic University. The first test showed that the trajectory of a moving user could be obtained with a standard deviation of about ±3–5 m. The main disadvantage of WiFi fingerprinting, however, is the required time consuming and costly signal strength system calibration in the beginning. Therefore, the authors have investigated if the measured signal strength values can be converted to the corresponding range to the access point. A new approach for this conversion is presented and analysed in typical test scenarios.     

通信·导航·物联一体化5G室内通信网络

4G 时代运营商已经沦为数据业务管道,5G 时代运营商将积极寻求业务转型。根据研究机构预测,室内定位及物联网市场前景广阔,运营商正在努力拓展相关领域的垂直行业应用。但是,现有室内通信网络功能单一,无法有效支撑室内定位和物联网业务。提出了一种通信·导航·物联一体化5G室内通信网络,通过在室分天线内部集成蓝牙模块,使其具备下行蓝牙定位、广告信息推送、链路损耗检测、上行物联收集和上行蓝牙定位等功能。该技术方案可应用于智慧医疗场景中,提供智慧导诊、设备管理、安防管理、后备保障和一键告警等服务。     

Extreme learning machine for indoor location fingerprinting

With the rapid growing market of wireless devices, positioning systems that make use of the signal strength of wireless devices are gaining more interest nowadays. Being able to track the location of a Wi-Fi or Radio Frequency Identification device could improve the quality of services in various sectors, including security, warehouse, logistic management, and healthcare. As compared with outdoor environment, positioning systems face a greater challenge in indoor environment because wireless signal is significantly influenced by building layout and surrounding objects, for which a location fingerprinting approach is needed. Moreover, the signal strength of a wireless device may also change over time, which is known as temporal variation, and therefore a reliable location estimation system must have the ability to learn and adapt with temporal changes. However, if the learning process is highly complex and requires long processing time, deploying the system into a larger scale would not be feasible. In recent years, Extreme Learning Machine (ELM) has surfaced as a viable alternative that challenged the norm of iterative and progressive learning. ELM has also been considered as a solution for indoor location fingerprinting. However, there has not been a comprehensive review on how the ELM-based approaches are linked with existing location fingerprinting techniques. Here we discuss some major location fingerprinting techniques, which are nearest-neighbor, LANDMARC, and LEMT, and formulate a new framework for systematically translating the techniques into ELM-based methods.     

基于地磁轨迹信号的新型室内定位

无处不在的地磁场由于室内环境中建筑结构的差异而具有独特的特征。此外,地磁信号的分辨难度会导致定位结果的不准确。本文提出了一种使用深度神经网络来提高定位精度的地磁室内定位系统。为了解决地磁场的低分辨率问题,本文将连续的地磁信号矢量化为轨迹序列,并以此为基础设计了一种新的地图构建方法来搭建用于室内定位的地磁数据库。然后,通过引入时间卷积网络(TCN)来提取磁轨迹序列的深层特征。实验结果表明,这种方法优于KNN和基于LSTM的DRNN等其他机器学习算法。      

物联网室内定位技术对比分析

随着物联网应用的发展,时空大数据对智慧城市等物联网应用具有决定性的意义。现在的物联网已经不是万物互联了,而是包含时空数据的、融合位置/时间/信息的事联网。人们的生活有70%的活动发生在室内,因此室内的定位就是物联网不可或缺的部分。详细分析了各种室内定位技术,并且对室内定位技术进行了对比分析,最后得出伪卫星是实现室内室外无缝定位的较好的方法。     

软件漏洞智能化挖掘技术研究进展

软件规模的不断扩大和新技术平台的发展对软件漏洞挖掘方法提出了新的挑战。在突破漏洞挖掘技术瓶颈的过程中,研究人员将机器学习方法应用于漏洞挖掘,利用机器学习模型自动学习代码的深层语法和语义规律,以提高漏洞挖掘的智能化水平和有效性,软件漏洞智能化挖掘技术已成为当前研究的热点。围绕软件漏洞智能化挖掘技术的研究展开分析,从静态挖掘和动态挖掘2个方面,对机器学习与漏洞挖掘技术结合的研究进行了深入分析。在漏洞智能化静态挖掘方面,从基于代码度量、基于代码模式和基于代码相似性3个方面梳理了现有研究工作;在漏洞智能化动态挖掘方面,则分类阐述了机器学习方法与动态挖掘方法结合的相关研究。依据对现有工作的总结,对未来漏洞智能化挖掘的发展趋势进行了展望。     

Application of meta-learning in cyberspace security: a survey

In recent years, machine learning has made great progress in intrusion detection, network protection, anomaly detection, and other issues in cyberspace. However, these traditional machine learning algorithms usually require a lot of data to learn and have a low recognition rate for unknown attacks. Among them, “one-shot learning”, “few-shot learning”, and “zero-shot learning” are challenges that cannot be ignored for traditional machine learning. The more intractable problem in cyberspace security is the changeable attack mode. When a new attack mode appears, there are few or even zero samples that can be learned. Meta-learning comes from imitating human problem-solving methods as humans can quickly learn unknown things based on their existing knowledge when learning. Its purpose is to quickly obtain a model with high accuracy and strong generalization through less data training. This article first divides the meta-learning model into five research directions based on different principles of use. They are model-based, metric-based, optimization-based, online-learning-based, or stacked ensemble-based. Then, the current problems in the field of cyberspace security are categorized into three branches: cyber security, information security, and artificial intelligence security according to different perspectives. Then, the application research results of various meta-learning models on these three branches are reviewed. At the same time, based on the characteristics of strong generalization, evolution, and scalability of meta-learning, we contrast and summarize its advantages in solving problems. Finally, the prospect of future deep application of meta-learning in the field of cyberspace security is summarized.     

高精度室内定位算法与技术综述

近年来高精度室内定位算法和技术的研究成为热点。任何一种孤立的定位技术由于其技术的独特性、局限性,无法满足现实需要。因此,本文对室内定位相关技术进行综合。首先对室内定位技术的应用前景做简单介绍,然后对目前主流的室内定位信道模型进行总结,然后对各种常用室内定位信标技术进行综述。最后指出采用单独的信标和定位技术,很难满足应用需求,多信标融合的室内定位技术将成为主流。