Enhanced Fingerprinting Based Indoor Positioning Using Machine Learning

Due to the inability of the Global Positioning System(GPS)signals to penetrate through surfaces like roofs,walls,and other objects in indoor environments,numerous alternative methods for user positioning have been presented.Amongst those,the Wi-Fi fingerprinting method has gained considerable interest in Indoor Positioning Systems(IPS)as the need for lineof-sight measurements is minimal,and it achieves better efficiency in even complex indoor environments.Offline and online are the two phases of the fingerprinting method.Many researchers have highlighted the problems in the offline phase as it deals with huge datasets and validation of Fingerprints without pre-processing of data becomes a concern.Machine learning is used for the model training in the offline phase while the locations are estimated in the online phase.Many researchers have considered the concerns in the offline phase as it deals with huge datasets and validation of Fingerprints becomes an issue.Machine learning algorithms are a natural solution for winnowing through large datasets and determining the significant fragments of information for localization,creating precise models to predict an indoor location.Large training sets are a key for obtaining better results in machine learning problems.Therefore,an existing WLAN fingerprinting-based multistory building location database has been used with 21049 samples including 19938 training and 1111 testing samples.The proposed model consists of mean and median filtering as pre-processing techniques applied to the database for enhancing the accuracy by mitigating the impact of environmental dispersion and investigated machine learning algorithms(kNN,WkNN,FSkNN,and SVM)for estimating the location.The proposed SVM with median filtering algorithm gives a reduced mean positioning error of 0.7959 m and an improved efficiency of 92.84%as compared to all variants of the proposed method for 108703 m^(2) area.

An Indoor Positioning Scheme for Visible Light Using Fingerprint Database with Multi.Parameters

This paper proposes a novel indoor positioning scheme based on visible light communication(VLC).A new indoor VLC positioning scheme using fingerprint database with multi-parameters have been raised.We conduct simulation and experimental research on the illumination intensity distribution of several direction parameters.In the experiment,four LED matrixes are identified by LED-ID with room dimensions of 3.75×4.00×2.7 m^3.The results show that the mean of the location error is 0.22 m in the receiving plane,verifying the correctness and feasibility of the positioning scheme.

5G Ultra-Dense Network Fingerprint Positioning Method Based on Matrix Completion

The problem of high-precision indoor positioning in the 5G era has attracted more and more attention.A fingerprint location method based on matrix completion(MC-FPL)is proposed for 5G ultradense networks to overcome the high costs of traditional fingerprint database construction and matching algorithms.First,a partial fingerprint database constructed and the accelerated proximal gradient algorithm is used to fill the partial fingerprint database to construct a full fingerprint database.Second,a fingerprint database division method based on the strongest received signal strength indicator is proposed,which divides the original fingerprint database into several sub-fingerprint databases.Finally,a classification weighted K-nearest neighbor fingerprint matching algorithm is proposed.The estimated coordinates of the point to be located can be obtained by fingerprint matching in a sub-fingerprint database.The simulation results show that the MC-FPL algorithm can reduce the complexity of database construction and fingerprint matching and has higher positioning accuracy compared with the traditional fingerprint algorithm.

Fingerprint Database Updating Using Crowdsourcing in Indoor Bluetooth Positioning System

Fingerprint⁃based Bluetooth positioning is a popular indoor positioning technology.However,the change of indoor environment and Bluetooth anchor locations has significant impact on signal distribution,which will result in the decline of positioning accuracy.The widespread extension of Bluetooth positioning is limited by the need of manual effort to collect the fingerprints with position labels for fingerprint database construction and updating.To address this problem,this paper presents an adaptive fingerprint database updating approach.First,the crowdsourced data including the Bluetooth Received Signal Strength(RSS)sequences and the speed and heading of the pedestrian were recorded.Second,the recorded crowdsourced data were fused by the Kalman Filtering(KF),and then fed into the trajectory validity analysis model with the purpose of assigning the unlabeled RSS data with position labels to generate candidate fingerprints.Third,after enough candidate fingerprints were obtained at each Reference Point(RP),the Density⁃based Spatial Clustering of Applications with Noise(DBSCAN)approach was conducted on both the original and the candidate fingerprints to filter out the fingerprints which had been identified as the noise,and then the mean of fingerprints in the cluster with the largest data volume was selected as the updated fingerprint of the corresponding RP.Finally,the extensive experimental results show that with the increase of the number of candidate fingerprints and update iterations,the fingerprint⁃based Bluetooth positioning accuracy can be effectively improved.

Optimization of MAC algorithm based on IEEE 802.15.4 in indoor positioning system

The mobility of the targets asks for high requirements of the locating speed in indoor positioning systems.The standard medium access control(MAC)algorithm will often cause lots of packet conflicts and high transmission delay if multiple users communicate with one beacon at the same time,which will severely limit the speed of the system.Therefore,an optimized MAC algorithm is proposed based on channel reservation to enable users to reserve beacons.A frame threshold is set to ensure the users with shorter data frames do not depend on the reservation mechanism,and multiple users can achieve packets switching with relative beacon in a fixed sequence by using frequency division multiplexing technology.The simulation results show that the optimized MAC algorithm proposed in this paper can improve the positioning speed significantly while maintaining the positioning accuracy.Moreover,the positioning accuracy can be increased to a certain extent if more channel resources can be obtained,so as to provide effective technical support for the location and tracking applications of indoor moving targets.

Research on Indoor Passive Positioning Technology Based on WiFi

In recent years,WiFi indoor positioning technology has become a hot research topic at home and abroad.However,at present,indoor positioning technology still has many problems in terms of practicability and stability,which seriously affects the accuracy of indoor positioning and increases the complexity of the calculation process.Aiming at the instability of RSS and the more complicated data processing,this paper proposes a low-frequency filtering method based on fast data convergence.Low-frequency filtering uses MATLAB for data fitting to filter out low-frequency data;data convergence combines the mean and multi-data parallel analysis process to achieve a good balance between data volume and system performance.At the same time,this paper combines the position fingerprint and the relative position method in the algorithm,which reduces the error on the algorithm system.The test results show that the strategy can meet the requirements of indoor passive positioning and avoid a large amount of data collection and processing,and the average positioning error is below 0.5 meters.

Multi-cluster-center based filtering algorithm and its application to WLAN indoor positioning

Wireless local area network(WLAN) is developing to a ubiquitous technique in daily life.As a related product,WLAN based indoor positioning system is attracting more and more concern.Fingerprint is a mainstream method of wireless indoor positioning.However,it still has some shortcomings of that received signal strength(RSS) is multi-modal and sensitive to environmental factors.These characters would have a negative effect on the performance of positioning system.In this paper,a filtering algorithm based on multi-cluster-center is proposed.We make full use of this algorithm to optimize the training samples at off-line phase to improve the performance of non-linear fitting with the fingerprint feature,and further enhance the positioning accuracy.Finally,we use multiple sets of original WLAN signal samples and signal samples after filtering as the training input of positioning system respectively.After that,the results analysis is demonstrated.Simulation results show that it is a reliable algorithm to enhance the performance of WLAN indoor positioning.

Coordinate correction algorithm for WLAN indoor positioning

In the fingerprint matching-based wireless local area network(WLAN) indoor positioning system,Kalman filter(KF) is usually applied after fingerprint matching algorithms to make positioning results more accurate and consecutive.But this method,like most methods in WLAN indoor positioning field,fails to consider and make use of users' moving speed information.In order to make the positioning results more accurate through using the users' moving speed information,a coordinate correction algorithm(CCA) is proposed in this paper.It predicts a reasonable range for positioning coordinates by using the moving speed information.If the real positioning coordinates are not in the predicted range,it means that the positioning coordinates are not reasonable to a moving user in indoor environment,so the proposed CCA is used to correct this kind of positioning coordinates.The simulation results prove that the positioning results by the CCA are more accurate than those calculated by the KF and the CCA is effective to improve the positioning performance.

基于FP-树的时空关联规则挖掘算法研究

提出一种基于FP-树的时空关联规则挖掘算法,即STFP-tree(Spatio-temporal FP-tree)算法.算法对时空数据进行时间上的划分,并用空间谓词分析空间对象间的位置关系,最终生成同时考虑时间和空间特征的FP-树,进而挖掘出强时空关联规则.实验结果证明了算法的有效性,适用于对时空数据进行分析挖掘.

HUID:DBN-Based Fingerprint Localization and Tracking System with Hybrid UWB and IMU

High-precision localization technology is attracting widespread attention in harsh indoor environments.In this paper,we present a fingerprint localization and tracking system to estimate the locations of the tag based on a deep belief network(DBN).In this system,we propose using coefficients as fingerprints to combine the ultra-wideband(UWB)and inertial measurement unit(IMU)estimation linearly,termed as a HUID system.In particular,the fingerprints are trained by a DBN and estimated by a radial basis function(RBF).However,UWB-based estimation via a trilateral method is severely affected by the non-line-of-sight(NLoS)problem,which limits the localization precision.To tackle this problem,we adopt the random forest classifier to identify line-of-sight(LoS)and NLoS conditions.Then,we adopt the random forest regressor to mitigate ranging errors based on the identification results for improving UWB localization precision.The experimental results show that the mean square error(MSE)of the localization error for the proposed HUID system reduces by 12.96%,50.16%,and 64.92%compared with that of the existing extended Kalman filter(EKF),single UWB,and single IMU estimation methods,respectively.

Robust Fingerprint Construction Based on Multiple Path Loss Model (M-PLM) for Indoor Localization

A robust radio map is essential in implementing a fingerprint-based indoor positioning system(IPS).However,the offline site survey to manually construct the radio map is time-consuming and labour-intensive.Various interpolation techniques have been proposed to infer the virtual fingerprints to reduce the time and effort required for offline site surveys.This paper presents a novel fingerprint interpolator using a multi-path loss model(MPLM)to create the virtual fingerprints from the collected sample data based on different signal paths from different access points(APs).Based on the historical signal data,the poor signal paths are identified using their standard deviations.The proposed method reduces the positioning errors by smoothing out the wireless signal fluctuations and stabilizing the signals for those poor signal paths.By consideringmultipath signal propagations from different APs,the inherent noise from these signal paths can be alleviated.Firstly,locations of the signal data with standard deviations higher than the threshold are identified.The new fingerprints are then generated at these locations based on the proposed M-PLM interpolation function to replace the old fingerprints.The proposed technique interpolates virtual fingerprints based on good signal paths with more stable signals to improve the positioning performance.Experimental results show that the proposed scheme enhances the positioning accuracy by up to 44%compared to the conventional interpolation techniques such as the Inverse DistanceWeighting,Kriging,and single Path LossModel.As a result,we can overcome the site survey problems for IPS by building an accurate radio map with more reliable signals to improve indoor positioning performance.

An Indoor Localization Approach Based on Fingerprint and Time-Difference of Arrival Fusion

In this paper,an effective target locating approach based on the fingerprint fusion posi-tioning(FFP)method is proposed which integrates the time-difference of arrival(TDOA)and the received signal strength according to the statistical variance of target position in the stationary 3D scenarios.The FFP method fuses the pedestrian dead reckoning(PDR)estimation to solve the moving target localization problem.We also introduce auxiliary parameters to estimate the target motion state.Subsequently,we can locate the static pedestrians and track the the moving target.For the case study,eight access stationary points are placed on a bookshelf and hypermarket;one target node is moving inside hypermarkets in 2D and 3D scenarios or stationary on the bookshelf.We compare the performance of our proposed method with existing localization algorithms such as k-nearest neighbor,weighted k-nearest neighbor,pure TDOA and fingerprinting combining Bayesian frameworks including the extended Kalman filter,unscented Kalman filter and particle fil-ter(PF).The proposed approach outperforms obviously the counterpart methodologies in terms of the root mean square error and the cumulative distribution function of localization errors,espe-cially in the 3D scenarios.Simulation results corroborate the effectiveness of our proposed approach.

面向无人智能小车的双验证安全定位方法

针对无人智能小车在网络、硬件、操作系统和软件方面存在众多安全隐患,易受到物理或远程安全攻击,使其偏离配送轨迹导致配送任务失败,甚至被攻击者操控干扰工厂正常运行的问题,提出了一种面向无人智能小车的双验证安全定位方法。在无人智能小车端,利用已有的Wi-Fi网络基础设施进行指纹定位,并设计特征融合策略实现Wi-Fi和磁场指纹的动态融合;在环境端,部署多个监测点采集无人智能小车发出的声音信号计算到达时间差,并根据空间分割方法计算小车位置。在此基础上,通过将无人智能小车上报的位置信息和监测点计算的位置坐标进行对比验证,一旦发现小车位置出现异常则进行异常告警,从而保证无人智能小车的正常运转工作。在真实室内场景下的实验结果表明,所提方法可以有效跟踪目标设备的位置坐标,定位精度优于现有基准算法。

一种基于SO-CNN模型的可见光室内定位优化方法

针对基于机器学习的可见光室内定位方法存在的手工调参、定位精度低等问题,结合蛇优化(Snake Optimization,SO)算法的寻优能力与卷积神经网络(Convolutional Neural Network,CNN)处理复杂非线性问题的能力,提出了一种基于SO-CNN模型的可见光室内定位优化方法。在考虑多径效应影响的情况下,采集每个位置点处的信噪比和对应位置坐标构建指纹数据库,对SO-CNN模型进行训练和测试,以得到最佳定位模型。实验结果表明,在5 m×5 m×3 m的房间中,与未经优化的CNN相比,该方法的平均定位误差降低了35.13%;与反向传播神经网络(Back Propagation Neural Network,BPNN)、多层感知器(Multilayer Perceptron,MLP)、SO-MLP相比,该方法的平均定位误差分别降低了54.75%,48.08%,37.01%。

基于深度学习的可见光通信系统室内三维定位

针对目前室内可见光通信系统三维定位的准确率与定位速度依然不佳的问题,提出一种基于深度学习的可见光通信系统室内定位方法。首先,设计了一个神经网络将指纹数据编码成二维阵列,利用卷积神经网络学习指纹阵列与目标位置之间的关系;然后,通过粒子群优化算法自动搜索卷积神经网络的超参数,以降低深度神经网络的训练难度。此外,设计了定位数据训练集、验证集与测试集的划分方法,有助于缓解神经网络的过拟合问题,并提高定位准确性。仿真结果表明,所提方法在6×6×4m3室内环境下的平均定位误差为0.024m,平均定位时间为0.478s。

基于位置信息指纹的蓝牙/WiFi混合定位方法

针对蓝牙/无线保真(WiFi)混合定位精度不理想、稳定性差等问题,提出了基于位置信息指纹的蓝牙/WiFi混合定位方法,该方法由离线阶段与在线阶段组成。在离线阶段,首先将采集的蓝牙/WiFi信号强度分为2组,第1组用于构建蓝牙、WiFi和蓝牙/WiFi混合指纹库;第2组作为训练指纹,分别与蓝牙、WiFi及蓝牙/WiFi混合指纹库匹配定位,以获得蓝牙、WiFi及蓝牙/WiFi混合指纹估计位置。随后,基于指纹估计位置和参考点构建位置信息指纹库。在在线阶段,先进行蓝牙、WiFi和蓝牙/WiFi混合指纹定位,然后结合蓝牙、WiFi和混合指纹估计位置生成在线位置信息指纹,最后,利用K近邻(KNN)算法实现与位置信息指纹库的匹配定位。实验结果表明,提出方法在2个公开数据集上的定位效果明显优于加权K近邻(WKNN)、高斯过程回归(GPR)和支持向量机(SVM)方法。在数据集一中,提出方法的均方根误差(RMSE)比WKNN、GPR和SVM最少减小了41.21%、48.33%和67.56%;在数据集二中,提出方法的平均绝对误差(MAE)为0.914 m,远优于WKNN、GPR和SVM。

面向室内语音导航的“附近”关系协同定位方法

针对传统室内定位模式单一,且单一“附近”空间关系无法满足定位需求的问题,提出一种融合“附近”空间关系/无线保真(WiFi)/行人航迹推算(PDR)/地磁的室内定位方法:阐述多源融合语音交互室内定位方法框架;并针对室内环境,建立“附近”空间关系概率密度函数;然后步行构建地磁/WiFi指纹库;最后后通过“附近”空间关系约束粒子及指纹匹配空间范围,在PDR推算位置基础上,利用WiFi/地磁/“附近”空间关系更新粒子权重,实现最优位置估计。实验结果表明,基于粒子滤波融合单一“附近”空间关系/WiFi/PDR/地磁为一种低成本可行的定位方法,平均定位精度1.93 m,有90%可能性定位精度可达2.52 m。

基于MVO-SVR的室内指纹定位算法

针对室内定位过程中由非视距和环境干扰导致的定位精度不高的问题,提出一种基于多元宇宙优化支持向量回归的室内指纹定位算法。首先通过基于超宽带通信技术的双边双向测距算法计算得到测距信息;然后利用测距值作为指纹特征并建立指纹库,使用SVR算法构建定位坐标和测距值之间的映射关系;最后使用MVO优化算法寻优SVR算法的cost和γ参数,以提升定位精度。实验显示,选择径向基函数作为SVR模型的核函数能够有效提高定位精度,并将MVO-SVR的结果与三边定位、随机森林算法、极致梯度提升算法、SVR的结果进行对比和分析,X方向平均绝对误差分别降低了20.12%、54.43%、60.66%和16.21%,Y方向平均绝对误差分别降低了79.57%、54.18%、59.29%和38.17%,平均定位误差Ep分别降低了60.73%、54.38%、60.01%和22.84%,且MVO-SVR算法在X和Y方向平均绝对误差均达到了厘米级。结果证明:基于MVO-SVR的室内指纹定位算法明显提升了定位精度,在复杂室内环境中具有良好的应用前景。

基于UHPLC-Q-Exactive Orbitrap MS的芪龙壮儿口服液指纹图谱研究

目的:建立芪龙壮儿口服液UHPLC-Q-Exactive Orbitrap MS正、负离子两种检测模式下的指纹图谱,为该制剂的质量控制提供参考和依据。方法:采用Thermo Accucore aQ RP18(2.1 mm×150 mm,2.6μm)色谱柱进行色谱分离,加热电喷雾离子源正、负离子同时检测。根据一级准分子离子精确质荷比、二级质谱碎片离子信息、文献报道和对照品比对鉴定共有峰成分。结果:构建了芪龙壮儿口服液正、负离子模式下的指纹图谱共有模式并分别确认了24(正离子模式)和17(负离子模式)个共有指纹峰。10批芪龙壮儿口服液指纹图谱与对照指纹图谱的相似度分别大于0.986(正离子模式)和0.968(负离子模式)。在正离子模式下鉴定了15个共有峰,其中9个共有峰采用对照品确认;负离子模式下鉴定了12个共有峰,其中9个共有峰采用对照品确认。结论:所建立的芪龙壮儿口服液正、负离子两种模式下的指纹图谱,结合共有峰的高分辨质谱定性分析,从正、负两个维度较为全面地表征了该制剂的化学成分指纹特征,为芪龙壮儿口服液的质量评价提供了依据和方法。

一种WiFi混合滤波的RSSI室内定位算法

针对接收信号强度指示(RSSI)在基于WiFi的室内定位应用中易受到环境影响出现波动、突变进而导致定位精度降低的问题,提出了一种WiFi混合滤波的RSSI室内定位算法。首先,根据WiFi信号距离衰减模型设计一种自适应阈值滤波器剔除观测信号异常值;然后,根据自适应阈值滤波、卡尔曼(Kalman)滤波、均值(Mean)滤波设计了一种混合滤波算法(ATKM)对RSSI数据进行滤波;最后,提出了一种基于斯皮尔曼-欧几里得距离的加权K近邻算法(SEWKNN)进行位置估计。在真实环境下的实测结果表明,本文提出的ATKM滤波算法能显著降低RSSI信号的波动,提出的SEWKNN算法在室内环境中平均定位误差为1.17 m,在走廊环境中平均定位误差为1.53m,相比传统的WKNN算法平均定位误差分别下降18.18%和16.84%。