基于RSS相关性的位置指纹室内定位方法

解决设备差异性造成的Wi-Fi信号强度不确定问题是位置指纹室内定位应用与推广的关键.一种基于设备间接收信号强度(Received Signal Strength,RSS)相关性的位置指纹室内定位方法被提出.以智能手机为用户终端,离线阶段,通过智能手机扫描的Wi-Fi信号强度信息,经过数据处理,筛选稳定的接入点(Access Point,AP),构建离线指纹数据库;在线定位阶段,对于实时获取的Wi-Fi信号强度信息,进行筛选处理后,挑选与离线指纹共同拥有的AP,并根据该AP集合,形成新的离线指纹和在线指纹.对离线指纹按RSS的大小降序排序;在线指纹,则以同一次序对RSS排序,然后利用皮尔逊相关系数和杰卡德相似系数,计算指纹相似度并排序,通过K最近邻(K-Nearest Neighbor,KNN)算法实现用户定位.实验表明该方法可有效解决设备差异性问题,并实现精确定位,平均定位误差达到1.7 m.     

基于Wi-Fi指纹和随机森林的定位算法

针对Wi-Fi信号强度的相似性对室内定位的影响,本文提出一种基于Wi-Fi指纹和随机森林的室内定位算法.该算法采用Wi-Fi作为信号源,以接收信号强度指示和基本服务集标识符来构建Wi-Fi指纹库,从而建立随机森林模型用于室内位置感知.仿真实验表明,该算法的定位误差约为2.26 m,与同类算法相比,在执行时间和定位精度上具有较好的优越性,算法精度提高约3.2％.     

基于随机森林算法的室内可见光指纹定位方法

为进一步提高动态目标室内可见光定位追踪系统性能，提出了一种基于随机森林（RF）算法的室内可见光指纹定位方法。利用发光二极管（LED）的光强信号作为特征构建指纹数据库，应用指纹库中的数据训练决策树，引入RF算法进行初始定位，再通过卡尔曼滤波对初始位置估计进行优化，从而获得更准确的定位轨迹。仿真结果表明：在5 m×5 m×3 m的室内场景下，通过所提定位方法能获得大部分采样点误差分布在4 cm之内的定位效果；此外，通过与不同室内可见光定位算法的性能进行对比，验证了所提算法的技术优势。     

基于Wi-Fi的动态室内定位方法

针对目前动态室内定位方法定位精度不足，研究基于Wi-Fi的动态室内定位方法。该方法利用传感器采集Wi-Fi的RSS(received signal strength，接收信号强度)指纹信号后，使用改进非均值滤波算法去除RSS指纹信号的干扰噪声，以不含干扰噪声的RSS指纹信号作为基础，使用基于指纹子空间匹配动态室内定位方法，计算不同阶段RSS指纹覆盖向量、汉明距离以及欧式距离等，得到若干个动态室内定位估计值，再使用峰值密度聚类算法对若干个动态室内定位估计值进行估计，获取估计值中可信的估计位置，即动态室内定位结果。实验结果表明：该方法不仅可有效去除RSS指纹信号含有的干扰噪声，还可对动态室内目标进行准确定位，定位误差仅为-1～0.5 m，定位精度较高。     

基于智能终端的室内定位系统研究与实现

针对全球定位系统GPS不能提供令人满意的室内定位结果,提出一种基于接收信号强度的WLAN指纹匹配定位技术,采用一种创新式的指纹库构建方式,改进加权K近邻算法,同时利用指纹匹配的优点来校准行人航位推算的累积误差,提高定位精度,设计完成一套基于智能终端的绝对定位系统。实验结果表明,与传统无线定位算法相比较,改进的无线指纹匹配定位平均误差为1.66 m,无线修正航位推算平均定位误差为0.56 m,达到了室内定位精度的标准,验证了改进算法的有效性及导航系统的实效性。     

一种基于位置指纹的室内WLAN定位方法

针对室内信号时变性导致定位不准的问题,提出了一种改进的3阶段位置指纹定位法。采样阶段,将采集信号的坐标、方位、接收信号强度的高斯分布及其对应的无线接入点等信息存储在数据库中生成位置指纹;在校正阶段中,利用参考点间信号强度的关联性信息,使用局部加权线性回归法,计算出一些虚拟点的信号强度;最后是线上实时定位阶段。通过与传统的加权K最邻近算法、直方图和联合聚类等3种定位方法相比较,该算法在同样的场景下可以取得更好的定位精度。     

基于差分信号强度的指纹室内定位算法

指纹定位以其低成本,高兼容性,高扩展性成为主流的室内定位技术。但是其依靠预先采集指纹强度的方式限定了其定位终端必须是采集信号的终端。这极大局限了指纹室内定位的应用场景。实验表明,不同终端之间在相同位置接收到相同接入点的信号可有5dBm-15dBm的误差。为了消去该误差对定位结果造成的影响,文中提出采用差分信号强度作为数据库指纹以计算定位结果。通过仿真,得到的结果表明这种定位方式明显提高了非采样终端的定位精度。     

室内定位系统中位置指纹库滤波算法研究

WiFi指纹定位是目前最受欢迎的室内定位技术之一,离线阶段创建的位置指纹库的精确与否对定位的精准度有很大的影响。传统的指纹库构建一般对采集的数据进行均值滤波,误差较大,并且在范围较大的场合进行定位时,由于位置指纹数据过多,使得定位时效性不好,结果不理想。为了提高指纹库的可靠性,通过给采集的源指纹数据赋予权值并根据权值的大小划分有效数据进行滤波处理,建立高精度指纹库,并利用基于临近域加权最近邻算法进行验证。实验结果表明,与传统的构建指纹库和定位算法相比,该方法显著提高了定位精度,缩短了定位时间。     

可见光重构指纹室内定位算法研究

提出一种可见光重构指纹室内定位算法(RFP),通过融合到达时间差(TDOA)算法和指纹算法可快速完成室内高精度定位。该算法首先利用TDOA算法多次估计得到的解集定义一个区域,然后在该区域中构建三维精细指纹库,再利用匹配算法定位未知节点。仿真结果表明,该算法的平均定位误差约为0.1719m,与传统的TDOA算法相比,提高了定位精度,与传统精细指纹算法相比,节省了定位时间。     

基于CSI相位矫正的室内指纹定位技术研究

针对室内定位服务的迫切需求,提出了基于信道状态信息相位作为指纹信息的室内定位方法,搭建了室内WiFi定位平台,实现了室内环境下的高精度位置估计。在离线阶段获取WiFi的信道状态信息,包括振幅、相位等参数,利用线性变换的方法对采集到的CSI相位信息进行预处理,建立鲁棒的信号指纹数据库。在线阶段提出了改进的加权KNN算法,对初次估计坐标进行二次匹配,从而完成设备及人员的定位。实验结果显示,平均定位精度达到了0.63 m,相较于传统的室内定位技术,定位精度有了明显的提升。     

Environment Adaptive Localization Method Using Wi-Fi and Bluetooth Low Energy

Received signal strength indicator (RSSI) based fingerprinting techniques for indoor positioning can be readily implemented via a wireless access point. These methods have therefore been widely studied in the field of positioning. However, fingerprinting suffers low accuracy of positioning on account of high noise occurrences which are caused by other wireless communication signals and environmental factors when the RSSI is received, and by relatively high errors on account of low position resolution compared to other methods such as time of flight and inertial navigation technology. In this paper, a modified fingerprint algorithm based on Wi-Fi and Bluetooth low energy applied to the log-distance path loss model is proposed to remove unnecessary Wi-Fi data, and produce the AP database that can be updated depending on the changes of the ambient environment as the indoor area is increasingly complicated and extended. Instead of using the existing fingerprinting techniques of consulting signal strengths as factors that are stored in a database, the proposed algorithm employs environmental variables to which the log-distance path loss model is applied. Therefore, the proposed algorithm has higher position resolution than existing fingerprint and can improve the accuracy of positioning because of its low dependence on reference points. To minimize database and eliminate inaccurate AP signals, the Hausdorff distance algorithm and median filter are applied. Using a database in which environment variables are stored, the results are inversely transformed into the log-distance path loss model for expression as coordinates. The proposed algorithm was compared with existing fingerprinting methods. The experimental results demonstrated the reduction of positioning improvement by 0.695 m from 2.758 to 2.063 m.     

基于指纹匹配的无蜂窝大规模MIMO三维定位方法

本文研究了无蜂窝大规模多输入多输出（Multiple input multiple output, MIMO）系统中基于指纹匹配的无线定位方法。假设服务区域内布设大量接入点（Access point, AP）,每个AP配置水平均匀线性阵列天线（Uniform linear array, ULA）或垂直ULA。利用相互正交的线性阵列天线（Orthogonal uniform linear array, O-ULA）对不同地理位置用户的方位角和俯仰角进行辨识,提取无线信道的角度功率谱矩阵构建方位角和俯仰角指纹库。借助谱聚类算法对指纹数据库进行预处理,然后通过两阶段指纹匹配策略计算指纹相似度并排序,在指纹库中搜索与用户指纹相似度最高的参考点,并利用加权K近邻算法（Weighted K-nearest neighbor, WKNN）估计用户位置。仿真结果表明,所提方案和单天线方案、ULA方案、均匀矩形阵列（Uniform rectangular array, URA）方案相比能够获得更高的三维定位精度。      

基于WiFi的自适应匹配预处理WKNN算法

针对基于接收信号强度（Received Signal Strength,RSS）的WiFi室内定位技术中,传统加权K邻近（Weighted K-nearest Neighbor,WKNN）算法不能自适应获取WLAN中有效接入点（Acess Point,AP）且参考点匹配准确度不高的问题,本文提出了自适应匹配预处理WKNN算法。该算法中每个实时定位点自适应地根据网络状况对AP的RSS均值由大到小排序,然后选择RSS均值较大的前M个AP,与参考点中对应的M个AP一起参与匹配预处理计算,从而优化了传统的指纹定位算法。同时将室内定位和室内地图相结合,使参考点和定位结果直观地展示在地图上,并通过使用地图数据大幅度简化了离线训练过程。此外,本文设计并实现了基于Android平台的室内定位系统,通过该系统验证了本文所提算法在单点定位和移动定位中的有效性。实验结果表明,该算法可获得30%以上的定位误差改善,有效提高了定位精度和定位稳定性。      

A Hierarchical Signal-Space Partitioning Technique for Indoor Positioning with WLAN to Support Location-Awareness in Mobile Map Services

The precise and accurate performance of location estimation is a vital component of context-aware applications. Numerous mobile devices with built-in IEEE 802.11 Wi-Fi technology can be used to estimate a user’s location through a wireless local area network (WLAN) in indoor environments in which fixed access points are deployed. This study deals with improving the common techniques of such positioning once the acquisition of the fingerprint database in offline phase is performed. The main idea is to propose a methodology that includes two layers of classification: a concurrent hierarchical partitioning of both signal and physical space in a way that signal patterns in each part of building have the highest similarity, and a precise and independent positioning in a given part. A procedure for combining the proposed classifier with either artificial neural network (ANN) or Bayesian probabilistic model is then introduced. We also consider an alternative strategy for ANN learning by including all raw observations. The average distance error was successfully reduced in the proposed methodology by 32 % compared to the simple approach. We concluded that the physical partitioning should also consider the signal behavior. Toosi location-aware mobile system was ultimately implemented, providing different services (e.g., friend finder and nearest point of interest) based on the proposed technique via WLAN. The system benefits from the high level of interaction provided by Asynchronous JavaScript and XML (AJAX) technology. It is capable of transferring locational data and GIS map services efficiently to the mobile terminal.     

融合反距离加权和矩阵填充的位置指纹库构建

为解决位置指纹定位在离线阶段构建位置指纹库时耗费的人力和时间成本较大,构建指纹库效率低和利用空间插值法构建的指纹库精度不高的问题,提出了一种融合反距离加权和矩阵填充的位置指纹库构建算法。该算法仅需人工采集定位区域内少量参考点的接收信号强度值用作信标点指纹信息,结合反距离加权算法特性计算出次信标点指纹信息,根据位置指纹库数据矩阵的低秩性,应用奇异值阈值矩阵填充算法构建出位置指纹数据库。仿真实验结果表明,所提算法有效降低了矩阵填充算法构建位置指纹库所需的人工和时间成本,构建出的位置指纹库定位性能优于反距离加权和克里金空间插值法,接近传统人工采集法,显著地提高了位置指纹库的构建效率。     

无线局域网室内定位关键参数计算

利用无线局域网（WiFi）进行定位是室内定位的一种可靠方法,但是在对城市WiFi信号进行采样收集,以利用它进行定位时会遇到许多现实困难.首先是安全隐私问题,往往无法进入他人住宅或办公区域测量WiFi接入点（AP）的实际位置;其次是终端WiFi模块的性能问题,不同WiFi模块在同一位置对同一AP可能获得不同的信号强度,因此也会造成信号强度偏移误差,最后是传播模型中的衰减参数估计问题,复杂的环境室内中往往难以靠经验确定信号衰减参数.文中提出一种基于Keenan-Motley （KM）模型的关键参数计算方法,通过对传播模型线性化求解能够利用周边采样点有效估算AP位置、信号强度偏移误差和传播衰减参数.     

Adaptive access points deployment for Indoor bluetooth positioning accuracy enhancement

The indoor positioning system based on fingerprint receives more and more attention due to its high positioning accuracy and time efficiency. In the existing positioning approaches, much consideration is given to the positioning accuracy improvement by using the angle of signal, but the optimization of access points (APs) deployment is ignored. In this circumstance, an adaptive APs deployment approach is proposed. First of all, the criterion of reference points (RPs) effective coverage is proposed, and the number of deployed APs in target environment is obtained by using the region partition algorithm and full coverage algorithm. Secondly, the wireless signal propagation model is established for target environment, and meanwhile based on the initial APs deployment, the simulation fingerprint database is constructed for the sake of establishing the discrimination function with respect to fingerprint database. Thirdly, the greedy algorithm is applied to optimize APs deployment. Finally, the extensive experiments show that the proposed approach is capable of achieving adaptive APs deployment as well as improving positioning accuracy.     

基于WiFi-GM指纹的室内定位算法

针对室内环境中单一指纹定位方法存在定位误差较大、定位漂移的问题,提出了一种融合室内Wi-Fi指纹和地磁指纹的定位算法。首先在大范围区域中通过K-means聚类方法将较大的匹配区域划分成更小的且特征更加明显集中的子区域,然后在在线阶段通过WiFi指纹粗定位到小区域,再通过地磁指纹定位系统进行近一步精匹配定位。实验表明,该融合算法缩小了地磁匹配的初始搜索范围,大大减少了指纹定位中的误匹配问题。实验中,平均定位误差仅2.17 m,最大定位误差3.61 m,较单一指纹定位系统性能均有大幅度提升,证明该定位方法具有一定的可行性与先进性。     

基于RSSI梯度的AP定位算法(英)

Recent rapid rise of indoor location based services for smartphones has further increased the importance of precise localization of Wi-Fi Access Point(AP).However,most existing AP localization algorithms either exhibit high errors or need specialized hardware in practical scenarios.In this paper,we propose a novel RSSI gradient-based AP localization algorithm.It consists of the following three major steps:firstly,it uses the local received signal strength variations to estimate the direction(minus gradient) of AP,then employs a direction clustering method to identify and filter measurement outliers,and finally adopts triangulation method to localize AP with the selected gradient directions.Experimental results demonstrate that the average localization error of our proposed algorithm is less than 2meters,far outperforming that of the weighted centroid approach.