Ultra-wideband (UWB) is renowned in Internet of Things (IoT) scenarios for its high positioning update rates and centimeter-level accuracy. Time Difference of Arrival (TDOA)-based positioning combines time measurements from multiple devices to estimate the location of a target device. However, these measurements are expressed in each device's own time scale, which is inherently offset and exhibits distinct time-varying drifts that must be compensated for during the estimation process. Inspired by this challenge, this paper introduces a novel sequential data-driven approach to wireless clock synchronization (WCS) based on the long short-term memory (LSTM). Experimental results show that, compared with the error accuracies of 0.0319 m and 0.0252 m relative to the two master anchors (MAs) achieved by traditional approaches relying on historical parameters to estimate future clock offsets, the proposed approach improves the accuracy to 0.0308 m and 0.0242 m. Furthermore, compared with deferred approaches that wait for the next Clock Correction Packet (CCP), the proposed approach achieves comparable estimation accuracy while reducing deferring from 0.1 s to 0.01 s. These results demonstrate that the proposed approach attains accurate tracking and prediction of clock deviations while maintaining real-time performance, achieving higher precision time mapping, and reducing TDOA measurement uncertainty.
Digital broadcasting signal is promised to be a positioning signal in indoors. A novel positioning technology named Time & Code Division-Orthogonal Frequency Division Multiplexing (TC-OFDM) is mainly discussed in this paper, which is based on China mobile multimedia broadcasting (CMMB). Signal strength is an important factor that affects the carrier loop performance of the TC-OFDM receiver. In the case of weak TC-OFDM signals, the current carrier loop algorithm has large residual carrier error. This paper proposes a novel carrier loop algorithm based on Maximum Likelihood Estimation (MLE) and Kalman Filter (KF) to solve the above problem. The discriminator of the current carrier loop is replaced by the MLE discriminator function in the proposed algorithm. The Levenberg-Marquardt (LM) algorithm is utilized to obtain the MLE cost function consisting of signal amplitude, residual carrier frequency and carrier phase, and the MLE discriminator function is derived from the corresponding MLE cost function. The KF is used to smooth the MLE discriminator function results, which takes the carrier phase estimation, the angular frequency estimation and the angular frequency rate as the state vector. Theoretical analysis and simulation results show that the proposed algorithm can improve the tracking sensitivity of the TC-OFDM receiver by taking full advantage of the characteristics of the carrier parameters. Compared with the current carrier loop algorithm, the tracking sensitivity is effectively improved by 3-5 dB in the simulation, and the better performance of the proposed algorithm is verified in the real environment.
In this paper, we propose UILoc, an unsupervised indoor localization scheme that uses a combination of smartphone sensors, iBeacons and Wi-Fi fingerprints for reliable and accurate indoor localization with zero labor cost. Firstly, compared with the fingerprint-based method, the UILoc system can build a fingerprint database automatically without any site survey and the database will be applied in the fingerprint localization algorithm. Secondly, since the initial position is vital to the system, UILoc will provide the basic location estimation through the pedestrian dead reckoning (PDR) method. To provide accurate initial localization, this paper proposes an initial localization module, a weighted fusion algorithm combined with a k-nearest neighbors (KNN) algorithm and a least squares algorithm. In UILoc, we have also designed a reliable model to reduce the landmark correction error. Experimental results show that the UILoc can provide accurate positioning, the average localization error is about 1.1 m in the steady state, and the maximum error is 2.77 m.
Indoor location-based service has widespread applications. With the ubiquitous deployment of WiFi systems, it is of significant interest to provide location-based service using standard WiFi devices. Most of the existing WiFi-based localization techniques are based on Received Signal Strength (RSS) measurements. As the bandwidth of WiFi systems increases, it is possible to achieve accurate timing-based positioning. This work presents a WiFi-based positioning system that has been developed at CSIRO as a research platform, where target devices are located using passive sniffers that measure the Time Difference of Arrival (TDOA) of the packets transmitted by the target devices. This work describes the architecture, hardware, and algorithms of the system, including the techniques used for clock synchronizing and system calibration. It is shown experimentally that the positioning error is 23 cm in open spaces and 1.5 m in an indoor office environment for a 80MHz WiFi system. The system can be used to track standard WiFi devices passively without interfering with the existing WiFi infrastructure, and is ideal for security applications.
Reference points (RP) clustering methods such as K-means are frequently used to reduce the region of search in most of fingerprint clustering algorithms. However, traditional clustering algorithms analysis the geometric proximity of RP only in the off line phase, which has nothing to do with the test point. Meanwhile, both the clustering pattern and the number of clusters need to be predefined directly or indirectly, which means an unsuitable clustering pattern or an unsuitable number of clusters would lead to poor estimation accuracy. In this letter, in order to improve the performance of KNN algorithm with the neighboring RP selection, we utilize the k-means clustering algorithm to analysis the geometric proximity between RP and test point in the online phase. In the proposed algorithm, k-means clustering algorithm groups M(M<K) nearest neighboring RPs according to their real physical distances to the test point instead of their signal distances. Then these M nearest neighboring RPs are used to estimate the location of the test point. Experiments were conducted in the fourteenth floor within an office building and the results demonstrate that the proposed method considerably outperforms the KNN algorithms in terms of positioning accuracy.
Babies have the ability to crawl, walk and climb between 6 months and 2 years of age, but they cannot speak languages. As a result, babies at this stage are unpredictable and their parents are unable to control their actions. Some places in the home are potentially dangerous to them. Parents cannot keep an eye on them all the time. There is a risk of injury when babies enter dangerous areas themselves. In this article, we propose iBaby, a low cost BLE pseudolite based indoor baby care system. Pseudolite, which consists of an MCU and multiple Bluetooth chips, along with navigation messages and an UKF based positioning engine, can achieve positioning accuracy better than 2 meters and a positioning response of 10 Hz. The system consists of pseudolite, server, client (parent and baby). It can run on the public mobile phone. Users do not need extra work, boot to use. Experiments show that the system can accurately locate the babies in the room and promptly remind their parents when the babies enter the dangerous area, which effectively reduces the possibility of the babies being injured.
A postures recognition and adaptive step detection algorithm based on low-cost microelectromechanical system (MEMS) sensors are proposed in this paper. The traditional Pedestrian Dead Reckoning (PDR) algorithm relies on the fixed sensor orientation, is unable to adapt to the high degree of freedom of hand-held terminal in navigation. The recognition method can recognize the postures of smartphone, such as texting, phoning, swinging and placed in the pocket, the algorithm consists of posture definition, feature extraction and class determination. An adaptive step detection algorithm is also proposed according to the detected device posture, the algorithm includes median threshold crossing detection, effective peak detection and adaptive step interval detection. The experiment results show that the proposed algorithm is able to identify the device postures effectively, and detect the pedestrian steps accurately. The accuracy of recognition can be above 95%, the accuracy of step detection can be above 97%. The research can provide important foundation for pedestrian dead reckoning based on hand-held terminals, and also be used in health management, behavior surveillance and other fields.
In the indoor environment, the activity of the pedestrian can reflect some semantic information. These activities can be used as the landmarks for indoor localization. In this paper, we propose a pedestrian activities recognition method based on a convolutional neural network. A new convolutional neural network has been designed to learn the proper features automatically. Experiments show that the proposed method achieves approximately 98% accuracy in about 2 s in identifying nine types of activities, including still, walk, upstairs, up elevator, up escalator, down elevator, down escalator, downstairs and turning. Moreover, we have built a pedestrian activity database, which contains more than 6 GB of data of accelerometers, magnetometers, gyroscopes and barometers collected with various types of smartphones. We will make it public to contribute to academic research.
Recently, many technologies such as Wi-Fi, camera, visible light, simultaneous localization and mapping(SLAM) are utilized to achieve high accuracy indoor localization. Using the visible light communication(VLC) technology, we can modulate light-emitting diode (LED) lights so that it can be adapted for localization as well as illumination, which is low-cost and practical. In this paper, we propose a new hybrid method, whose principle is using RGB-D camera to recognize the characteristics of the circumstance as the assistance of VLC positioning. The computer vision technology is employed to process the information captured by the RGB-D camera to help VLC positioning system derive a high accuracy location. Experiment results show that our sensor fusion positioning system is computationally efficient, and it can achieve one centimeter precision.
An indoor positioning method is proposed and evaluated. The method combines location fingerprinting and dead reckoning differently from conventional combinations. It utilizes a compound location fingerprint, which is composed of radio fingerprints at multiple points of time, that is, at multiple positions, and displacements between them estimated by dead reckoning. To avoid accumulated errors from dead reckoning, the method uses short-range dead reckoning. The method was evaluated in a student room whose size was 11 × 5 m and had furniture. Six Bluetooth beacons were placed in the room. The received signal strength indicator (RSSI) values of the beacons were collected at 28 measuring points, which were points of intersection on a 1-m by 1-m grid where no obstacles existed. A compound location fingerprint is composed of RSSI vectors at two points and a displacement vector between them. A support vector machine (SVM) and random forests (RF) were used to build regression models. The root mean square error (RMSE) of position estimation with SVM and RF was respectively 2.35 and 0.80 m. These errors were lower than those with a single-point baseline model, where a feature vector is composed of only RSSI values at one location. The results suggest that the proposed method is effective in indoor environments. We also discuss the relationship between the permissible error tolerance and cumulative percentages of correct answers, and the influence of dead reckoning errors on RMSE.
In traditional UWB (Ultra-wideband) indoor Iocalization algorithms, motion information and position changing of the target within a set of observation sequences is generally ignored, which makes it difficult to achieve high-precision positioning for dynamic targets, especially for high-speed moving objects. In this paper, a high-precision dynamic UWB indoor Iocalization algorithm considering speed parameters is proposed. This algorithm introduces speed parameters of the target to estimate its position changing within a set of observation sequences when establishing observation equations. Also, the virtual observations and prior constraints about the speed parameters are utilized to resolve the rank deficiency problem resulted by introducing speed parameters. The experimental results show that the proposed algorithm can make full use of the prior knowledge of the target motion and achieve real-time, stable and precise Iocalization for moving targets, without increasing any hardware cost.
The demand for location-based services (LBS) in large indoor spaces, such as airports, shopping malls, museums and libraries, has been increasing in recent years. However, there is still no fully applicable solution for indoor positioning and navigation like Global Navigation Satellite System (GNSS) solutions in outdoor environments. Positioning in indoor scenes by using smartphone cameras has its own advantages: no additional needed infrastructure, low cost and a large potential market due to the popularity of smartphones, etc. However, existing methods or systems based on smartphone cameras and visual algorithms have their own limitations when implemented in relatively large indoor spaces. To deal with this problem, we designed an indoor positioning system to locate users in large indoor scenes. The system uses common static objects as references, e.g., doors and windows, to locate users. By using smartphone cameras, our proposed system is able to detect static objects in large indoor spaces and then calculate the smartphones' position to locate users. The system integrates algorithms of deep learning and computer vision. Its cost is low because it does not require additional infrastructure. Experiments in an art museum with a complicated visual environment suggest that this method is able to achieve positioning accuracy within 1 m.
Magnetic in door positioning has aroused much attention in the past few years. However, its performance varies not only with available magnetic anomaly dataset but also with the quality of magnetometer measurements. As a magnetic instrument, the performance of a magnetometer is expressed by quite some indexes that are confusing. In this work, we propose a unified model in understanding the performance of a magnetometer. Based the model, we evaluated and compared the performance of popular magnetometers for in-door positioning. Our work is insightful in understanding the performance and limitations of magnetometers for those who want to study magnetic in-door positioning.
Indoor positioning system based on MEMS (Micro-electromechanical Systems) technology is really self-measurement technology which does not depend on any external signal and other infrastructure. Inertial/magnetic sensors based on MEMS technology are typically composed of three orthogonal gyroscopes, three orthogonal accelerometers and three orthogonal magnetometers, which can form AHRS (attitude heading reference system) and pedestrian dead reckoning system. However, the inertial/magnetic sensors can offer good short term positioning in indoor environments, but the absolute position fixes of medium-to long-term have always been the research hotspot and difficulty. Therefore, a hierarchical calibration architecture based on inertial/magnetic sensors is proposed creatively, including the data layer - a axial-based temperature variation model is proposed to calibrate the raw data from MEMS sensors by the analysis and modeling of sensor error characteristics; the signal layer - a triple zero velocity update method is proposed to update integral initial value according to the pedestrians' gait characteristic; the information layer - an intelligent information fusion method is proposed based on machine learning to calibrate the orientation and positioning. Besides, a 3-D indoor positioning platform is set up based on MPU9250, and the effectiveness of the proposed hierarchical calibration architecture is verified based on the platform.
Indoor localization based on Receive Signal Strength Indicator (RSSI) is widely used due to its low infrastructure cost. Unfortunately, RSSI is always fluctuating because of multipath effect, human movement and environmental change, which makes traditional methods get unexpected results. What's more, state-of-the-art clustering algorithm divides offline radio map into the different non-overlapping group while ignores the fact that different groups should share the same members. In this paper, a novel Robust Sparse Overlapping Group Lasso (RSOGL) algorithm is proposed for indoor localization. The scheme first utilizes similarity between the offline Reference Point (RP) and online Test Point (TP) to obtain overlapping group via Fuzzy C-means (FCM) clustering, and then uses RSOGSL algorithm to reconstruct TP's fingerprint. Our RSOGL system has been operated in a real environment. Experimental results demonstrate that proposed system outperforms traditional fingerprinting methods.
With the rapid development of national economy, higher requests have been put forward on indoor and outdoor seamless PNT (positioning navigation and timing) service. Meanwhile, with the popularity of smartphones and the rapid development of MEMS (microelectromechanical system) sensors, researches on smartphone-based high-precision, high-reliability, but low-cost seamless indoor/outdoor navigation method have become the highlighted topics. On the basis of the detailed analysis of smartphone camera's capability, this paper analyzed the method of coded targets aided pedestrian navigation techniques. With digitalized BIM (building information model), this paper proposed an algorithm for fast, precisely and robustly recognizing and matching artificial targets, from which the position and orientation of camera can be deduced and helps to improve the locating performance of smartphone. Laboratory experiments have been carried out to verify the feasibility of this vision-aided method: the locating precision of the proposed algorithm can reach cm level, whose exposure time is 50us and its corresponding process time is about 33ms, which could be of great use for the seamless indoor and outdoor locating with off-the-shelf smartphone.
Recent years, the use of widely covered Wi-Fi signal to achieve accurate indoor positioning has become a research hotspot. The Wi-Fi Positioning Algorithm based on the existing RF Fingerprint can obtain high positioning accuracy, but there is a big shortage of large deployment and maintenance cost. In order to reduce the cost of Wi-Fi fingerprints collection, this paper utilizes the RSS collected by the consumers using the mobile phone for electronic payment, constructs the low cost crowdsourcing Wi-Fi RF Fingerprint, and realizes the shop-level accurate positioning with CNN. Based on the feature of CNN extraction from local to global, this paper constructs a characteristic group which includes signal intensity, user transaction time and shop information. The statistics of each feature in the statistical interval are obtained by using the interval of 1 to 23 days before the current time. The Min-Max normalization of the statistical value is to avoid inconsistencies in the data distribution caused by the loss of data. In this way, the feature map of window length and statistic feature is constructed, the different feature groups of Wi-Fi and Shop are used as input matrices of multiple CNN, then the other manual features are combined to train a CNN positioning classifier to realize the position estimation of shop-level. A large number of experimental data tests show that the proposed algorithm can obtain higher positioning accuracy by 91% compared with the positioning algorithm based on LR(Logistic Regression), AdaBoost and XGBoost.
Currently, indoor navigation has become a hotspot in the field of Location Based Service. However, there is no universal model to geocode and locate interior facilities. In this paper, the indoor space is constructed of three layers, including venue, room, and POI. Then a geocoding framework is proposed to encode each POI as a code including three parts, namely partition number, functional number and connectivity number. In contrast to most indoor navigation methods which use WIFI or RFID would necessitate extra cost of money, manpower, and maintenance, we take advantages of QR code, which is money saving and convenient, to restore the location information. Taking the library of Wuhan University as a sample area, a prototype implementing our method for indoor navigation is generated. On this basis, a mobile navigation electronic map is developed and implemented, which can provide services of indoor positioning and optimal path push in the library.
In recent years, indoor positioning has been getting a lot of attention. An indoor location method based on optimal dilution of precision of displacement vector components and factor of weighting adjustment with multiple array pseudolites is proposed to solve two problems that: 1) a single array pseudolite drops sharply with the distance increasing between the receiver and pseudolite antennas; 2) the carrier phase measurement error increases and positioning accuracy decreases because of array psuedolite antennas aging or damage. In order to solve the first question, the dilution of precision of receiver displacement vector components are calculated in Newton-Raphson based on multiple array pseudolites, and the displacement value corresponding to x-coordinate (likewise y- and z-coordinate) minimum dilution of precision is involved in iterative calculation. In order to solve the second question, the idea of weighting factor adjustment is introduced in the Newton-Raphson. The theoretical analysis and simulation experiments show that this method can effectively improve indoor positioning accuracy and robustness.
Ultra-wide Band (UWB) plays an increasingly important role in indoor positioning due to its advantages of low emission power, high multipath resolution, and centimeter level positioning accuracy. However, the UWB system is vulnerable to Non-Line of Sight (NLOS) because of human or vehicle occlusion. In this paper, a novel Pedestrian Dead Reckoning (PDR) aided UWB indoor positioning method is proposed. Firstly, an NLOS accuracy factor model is constructed according to the distance difference between UWB and PDR. This accuracy factor judges whether UWB works normally or not. Once the accuracy factor exceeds a certain threshold, PDR positioning is adopted. Meanwhile, the direction and the position predicted by the extended extrapolation method from UWB are applied to provide the initial value for PDR processing. If the accuracy factor is within the threshold range, PDR is replaced by UWB. The experimental results show that the method proposed improves the positioning accuracy of UWB, and that the generalized extrapolation model can calculate more reliable initial values before PDR operates. The obtained results have validated that the proposed method can effectively restrain the impact of NLOS on stable and accurate positioning.
