Conference Paper

Lightweight map matching for indoor localisation using conditional random fields

Conference Paper

Lightweight map matching for indoor localisation using conditional random fields

If you want to read the PDF, try requesting it from the authors.

Abstract

Indoor tracking and navigation is a fundamental need for pervasive and context-aware smartphone applications. Although indoor maps are becoming increasingly available, there is no practical and reliable indoor map matching solution available at present. We present MapCraft, a novel, robust and responsive technique that is extremely computationally efficient (running in under 10 ms on an Android smartphone), does not require training in different sites, and tracks well even when presented with very noisy sensor data. Key to our approach is expressing the tracking problem as a conditional random field (CRF), a technique which has had great success in areas such as natural language processing, but has yet to be considered for indoor tracking. Unlike directed graphical models like Hidden Markov Models, CRFs capture arbitrary constraints that express how well observations support state transitions, given map constraints. Extensive experiments in multiple sites show how MapCraft outperforms state-of-the art approaches, demonstrating excellent tracking error and accurate reconstruction of tortuous trajectories with zero training effort. As proof of its robustness, we also demonstrate how it is able to accurately track the position of a user from accelerometer and magnetometer measurements only (i.e. gyro- and WiFi-free). We believe that such an energy-efficient approach will enable always-on background localisation, enabling a new era of location-aware applications to be developed.

No full-text available

Request Full-text Paper PDF

To read the full-text of this research,
you can request a copy directly from the authors.

... Magnetic field distortions are specific to building infrastructures and provide additional cues for the fingerprint-based system [16]. Map-Matching algorithms align motion trajectories with a floorplan via conditional random fields [17], hidden Markov models [18], [19], particle filtering [20], or dynamic programming [17]. The paper proposes a novel fusion of WiFi, IMU, and floorplan data by a combination of optimization and a convolutional neural network (CNN). ...
... Magnetic field distortions are specific to building infrastructures and provide additional cues for the fingerprint-based system [16]. Map-Matching algorithms align motion trajectories with a floorplan via conditional random fields [17], hidden Markov models [18], [19], particle filtering [20], or dynamic programming [17]. The paper proposes a novel fusion of WiFi, IMU, and floorplan data by a combination of optimization and a convolutional neural network (CNN). ...
... • Mapcraft is a state-of-the-art map-matching system [17]. 2 Our test-sites are shopping malls, where the dataset contains 3 buildings with 2 floorplan images each, where we create four testing setups for assessing the generalization capability over different building types. ...
Preprint
Full-text available
The paper proposes a multi-modal sensor fusion algorithm that fuses WiFi, IMU, and floorplan information to infer an accurate and dense location history in indoor environments. The algorithm uses 1) an inertial navigation algorithm to estimate a relative motion trajectory from IMU sensor data; 2) a WiFi-based localization API in industry to obtain positional constraints and geo-localize the trajectory; and 3) a convolutional neural network to refine the location history to be consistent with the floorplan. We have developed a data acquisition app to build a new dataset with WiFi, IMU, and floorplan data with ground-truth positions at 4 university buildings and 3 shopping malls. Our qualitative and quantitative evaluations demonstrate that the proposed system is able to produce twice as accurate and a few orders of magnitude denser location history than the current standard, while requiring minimal additional energy consumption. We will publicly share our code, data and models.
... Background and Related Works 15 ...
... Map pre-processing takes a floor plan as input, and produces a graph that encodes a set of discrete states (locations) and represents physical constraints between discrete states imposed by the map. This information will then be fed to the second step to define the feature functions then goes to the step to determine the feature weights and finally, system is ready now for computing user's position [15]. ...
... Most of the work done in this field adopted the actual floor map image to locate and display user position on the map [14], [15], [17]. Using actual floor map necessitates loading, in the smartphone subject to positioning, the map image of each floor in the building. ...
Thesis
Full-text available
The inspiration of this work is to build hybrid indoor positioning and monitoring system that integrates smartphone sensors and Wi-Fi fingerprinting to achieve reliable, flexible and accurate indoor positioning system. The most influential factor that is affected by such a system is the presenting of the floor real map to the system which is very tedious, time consuming, complex processing task to be presented on a smartphone. This thesis has suggested and adopted an idea of virtual map which is generated automatically inside the smartphone. Then a mechanism that combines the APs fingerprint together with smartphone sensors, accelerometer, magnetometer and barometer readings, has been created to accurately determine the user position inside the floor map relative to well-known landmarks in the floor. In addition, the proposed mechanism lets the user watching her/himself moving on the virtual map that reflects the real floor map. Also the proposed system offers a monitoring activity which lets the administrator to watch and locate certain user inside the building. The proposed system consists of two phases; the training phase which is used to collect the strongest signal strength, for each AP, in each floor in the building to be sent and stored in the database server. The localization phase is used to retrieve the strongest signal strength from the database server then in conjunction with current smartphone readings of AP signal strength, accelerometer, magnetometer and barometer are used to determine user position. The system was tested in two different indoor environments and achieved positioning accuracies of approximately 2 meters. Standard performance metrics have been used to compare the proposed system with latest existing systems. The proposed system is implemented on Galaxy S4 smartphone with Android platform using java programing language.
... When applying a fingerprinting method, a solution to optimize localization accuracy is sensor fusion, i.e., where data from other sensors is correlated with the RSS values. In the state of the art, sensor fusion methods such as Kalman filters [20] and conditional random fields [32] have been applied (see Section 7). In our IWK evaluation case study, the available mobile platforms (1 st gen iPad Mini; latest is iPad 6 th gen) only provided low-quality motion data, offered limited computational ability (1 GHz dual-core ARM, 512 MB) and had to run heavyweight apps concurrently (i.e., the iCare Adventure mobile game). ...
... Indoor maps are key to improving indoor localization accuracy as they allow reconciling observations, typically related to motion [10,32,52,56,58,59] but also RSS [60], with map constraints. When relying on Particle Filters (PF), indoor maps can be utilized to filter out erroneous particles (e.g., passing through walls) [10,52,56,58,59]. ...
... When relying on Particle Filters (PF), indoor maps can be utilized to filter out erroneous particles (e.g., passing through walls) [10,52,56,58,59]. For Conditional Random Fields (CFR), map constraints can be incorporated into feature functions to establish the degree to which observations support states and transitions [32]. ...
Article
Full-text available
In a digitally enabled healthcare setting, we posit that an individual’s current location is pivotal for supporting many virtual care services—such as tailoring educational content towards an individual’s current location, and, hence, current stage in an acute care process; improving activity recognition for supporting self-management in a home-based setting; and guiding individuals with cognitive decline through daily activities in their home. However, unobtrusively estimating an individual’s indoor location in real-world care settings is still a challenging problem. Moreover, the needs of location-specific care interventions go beyond absolute coordinates and require the individual’s discrete semantic location; i.e., it is the concrete type of an individual’s location (e.g., exam vs. waiting room; bathroom vs. kitchen) that will drive the tailoring of educational content or recognition of activities. We utilized Machine Learning methods to accurately identify an individual’s discrete location, together with knowledge-based models and tools to supply the associated semantics of identified locations. We considered clustering solutions to improve localization accuracy at the expense of granularity; and investigate sensor fusion-based heuristics to rule out false location estimates. We present an AI-driven indoor localization approach that integrates both data-driven and knowledge-based processes and artifacts. We illustrate the application of our approach in two compelling healthcare use cases, and empirically validated our localization approach at the emergency unit of a large Canadian pediatric hospital.
... Algizawy et al. [20] used HMM to generate a road-level traffic density, at an hourly granularity, for each mobile trajectory. Xiao et al. [21] used contextual relationships between trajectory points as features of the CDR trajectories in a conditional random field model to reconstruct individual trajectories. Chen et al. [22] proposed two approaches for completing CDRs adaptively to reduce the sparsity and mitigate the problems the latter raises. ...
... ext is the search space on R for T. p i to find matched R.q j . e entries of D are gradually filled as the dynamic programming proceeds (lines 4-10), and the last entry stores the LCSS of aligning T and R. Finally, we decode D to find all the aligning cell-id log pairs of the optimal alignment (lines [12][13][14][15][16][17][18][19][20][21][22]. e extended LCSS model can be viewed as a modified version of the models [26,27], which not only finds the longest common subsequence in terms of the accumulate number of matched anchors LCSS(T, R) but also the cycle numbers of the cell-id trajectory. ...
Article
Full-text available
Mobile phone data have become a critical data source for transportation research. While a cell-id trajectory was routinely reorganized by International Mobile Subscriber Identity (IMSI), it potentially allows to analyze transportation behaviors and social interaction of total population, with a full temporal coverage at low cost. However, cell-id trajectory is often sparse due to low reporting frequency and uncertainness of mobile holders’ position. So, the cell-id trajectory refinement has been recognized as challenging work to further facilitate trajectory data mining. This paper presents a comprehensive approach to identify cell-id trajectories of public service vehicles (PSVs) from large volume of trajectories and further refines these cell-id trajectories by a heuristic global optimization approach. The modified longest common subsequence (LCSS) method is used to match a cell-id trajectory and a public transportation route (PTR) and correspondingly calculates their similarities for determining whether the trajectory is PSV mode or not. Taking full advantages of the nature of a PSV tends to move on the PTR in uniform motion to meet a prescript visit to stops, a heuristic global optimization approach is deployed to build a spatiotemporal model of a PSV motion, which estimates new locations of cell-id trajectories on the PTR. The approach was finally tested using Beijing cellular network signaling datasets. The precision of PSV trajectory detection is 90%, and the recall is 88%. Evaluated by our GNSS-logged trajectories, the mean absolute error (MAE) of refined PSV trajectories is 144.5 m and the standard deviation (St. Dev) is 81.8 m. It shows a significant improvement in comparison of traditional interpolation methods.
... Both industry and academia currently pay much attention to indoor location based services (LBSs) and make every effort to improve positioning accuracy. Many novel methods have been developed; for example, some indoor localization applications locate users by using their relative locations to environmental physical features [1], [2]. WiFi and Bluetoothbased localization systems offer ubiquitous localization [3]- [7]. ...
... ViVi [34] Fingerprint spatial gradient, pattern matching Accurate, robust to temporal instability Affected by long-term changes Guo et al. [35] Windowing and sliding techniques based on a group of fingerprints, a fusion algorithm called multiple classifiers multiple samples are applicable for large-area buildings, such as a shopping mall and an engineering building, because they cannot detect enough landmarks with high-precision in residences due to a lack of specific places. Some systems leverage environmental physical features to conduct indoor localization without relying on the RF signature [1], [2], [41]- [43]. In such systems, users' relative positions to physical features and specific reference points are obtained from videos and images. ...
Article
Full-text available
Currently, indoor localization technology and indoor location-based services are becoming increasingly important in the area of mobile and ubiquitous computing. However, the design of an indoor location-based system confronts two challenges: achieving high-precision location recognition and identifying what indoor objects actually are (which is called semantic labeling). In this paper, we propose AtLAS, an activity-based indoor localization and semantic labeling mechanism. The key idea is that some objects in an indoor environment, such as doors and toilets, determine predictable human behaviors in small areas, which can be reflected in unique sensor readings. AtLAS leverages this idea to determine a user’s accurate location by identifying users’ activities. Furthermore, we leverage the topological structure of indoor objects to mine the semantic knowledge and label the objects through gained knowledge automatically. To the best of our knowledge, AtLAS is the first attempt to build a system that leverages users’ activities to conduct a high-precision indoor localization and semantic labeling system for the case of residences. Experimental results show that AtLAS can achieve a median localization accuracy of 0.57 meters, and the system can localize the landmarks with a median accuracy of 0.43 meters on average without 5% worst errors. AtLAS can label the objects semantically with a 5.7% false positive rate and a 5.8% false negative rate on average.
... funkbasierten Sensoren, wie z. B. Wi-Fi [394,188], GPS [223] oder Ultra-Breitband (UWB) [147], oder Magnetometer [178,460] oder mit Informationen über die Umgebung [61,232,479], da die technischen Herausforderungen wie der Sensordrift und die Zerlegung der Beschleunigung immer noch ungelöst sind. Auch rotationsinvariante Methoden, die frequenzbasierte Parameter verwenden, leiden dennoch noch unter einer schlechten Genauigkeit [105,52]. ...
Thesis
Full-text available
Standortbezogene Unterhaltung ist mittlerweile zu einem Grundbedürfnis geworden. Die erforderliche Genauigkeit und Zuverlässigkeit von Lokalisierungssystemen wächst nicht nur für intelligente Systeme wie selbstfahrende Fahrzeuge, Lieferdrohnen und mobile Geräte, sondern auch für alltägliche Fußgänger. Aufgrund der allgegenwärtigen Sensoren wie Kameras, GPS und Trägheitssensoren werden mit aufwendig handgefertigten Modellen und Algorithmen eine Vielzahl von Lokalisierungssystemen entwickelt. Um eine Einschränkung der freien Sicht und unterschiedliche Lichtverhältnisse von Kamerasystemen zu vermeiden werden typischerweise Funk- und Trägheitssensoren zur Lokalisierung verwendet. Unter idealen Laborbedingungen können diese Sensoren und Modelle, Positionen und Orientierungen langfristig genau abschätzen. In realen Umgebungen wirken sich jedoch viele Probleme wie ungenaue Systemmodellierung, unvollständige Sensormessungen, Rauschen und komplexe Umgebungsdynamiken auf die Genauigkeit und Zuverlässigkeit aus. Individuell betrachtet haben Funk- und Trägheitssensoren Schwierigkeiten: Funk lokalisiert aufgrund mehrerer Pfade durch statische oder dynamische Objekte entlang der Ausbreitungspfade zwischen Sender und Empfänger sehr ungenau. Im Gegensatz dazu akkumulieren Trägheitssensoren im Laufe der Zeit Entfernungs- und Orientierungsfehler und können keinen absoluten Bezug zur Weltkarte herstellen. Verfahren des Stands der Technik ergänzen beide Sensoren, um komplementäre Effekte zu verwenden, können jedoch die Schwierigkeiten nicht beheben. Darüber hinaus können sie mit einfachen Bewegungsmodellen wie konstanter Beschleunigung oder Geschwindigkeit keine stark nichtlinearen menschlichen Bewegungen beschreiben. Das Hauptziel dieser Arbeit ist es daher, die Auswirkungen datengetriebener Methoden und verschiedener Sensordatenströme von lose platzierten Sensoren auf die Genauigkeit der Schätzung menschlicher Posen in hochdynamischen Situationen zu untersuchen. Die absolute Genauigkeit der erhaltenen Ergebnisse wird mit Filtermethoden nach dem Stand der Technik verglichen. Um die Probleme von Menschen entworfenen Lokalisierungsmodellen zu lösen, werden in dieser Arbeit maschinelle und tiefe Lernmethoden verwendet. Es werden Lernmethoden zur Positions‐, Geschwindigkeits- und Orientierungsschätzung sowie zur Rekonstruktion der Trajektorie unter Verwendung multimodaler Messungen von Funk- und Trägheitssensoren vorgestellt, um eine genaue und robuste Lokalisierung zu erreichen. Die Auswirkungen datengetriebener Verfahren entlang einer typischen Verarbeitungskette für die Lokalisierung mit Funk- und Trägheitssensoren werden untersucht. Die Verarbeitungskette ist lose gekoppelt in atomare Komponenten unterteilt, sodass jedes datengetriebene Verfahren problemlos ausgetauscht werden kann. Sequenzbasierte Lernmethoden werden entlang der Verarbeitungskette verwendet, um absolute Positionen aus Ankunftszeitstempeln von Funksignalen mit Mehrwegeausbreitung zu schätzen, ungerichtete Geschwindigkeitsvektoren von Trägheitssensoren zu schätzen, Bewegungsmuster zu klassifizieren, die die Ausrichtung der Trajektorie kalibrieren und um schließlich die einzelnen Komponenten zu einer Trajektorie zu fusionieren. Die vorgeschlagenen Methoden lernen, mit unterschiedlichem Bewegungsverhalten umzugehen und ermöglichen eine robuste und präzise Lokalisierung. Im Rahmen von Großstudien werden Mess- und Referenzdaten mit verschiedenen Bewegungsformen bei unterschiedlichen Geschwindigkeiten erfasst. Umfangreiche Experimente zeigen die Wirksamkeit und das Potenzial der vorgeschlagenen Methoden. Die datengetriebene, modulare Verarbeitungskette liefert genauere und robustere Schätzungen als bekannte Verfahren, auch bei dynamischen Bewegungen mit verrauschten Trägheitssensoren und Funkumgebungen mit Mehrwegeausbreitung.
... As the proposed method is landmark-based, we draw comparisons to three competing state-of-the-art approaches for indoor localization with landmark-based magnetism mechanisms are covered in Ref. 3. Wang et al. 8 presented UnLoc, established on landmarks matching. MapCraft, introduced by Xiao et al., 9 is rooted on the conditional random fields method. The IODetector from Li et al. 10 used joint thresholds. ...
Article
Full-text available
Smartphone-based indoor localization methods are frequently employed for position estimation of users inside enclosures like malls, conferences, and crowded venues. Existing solutions extensively use wireless technologies, like Wi-Fi, RFID, and magnetic sensing. However, these approaches depend on the presence of active beacons and suitable mapping surveys of the deployed areas, which render them highly sensitive to the local ambient field clutters. Thus, current localization systems often underperform. We embed small-volume and large-moment magnets in pre-known locations and arrange them in specific geometric forms. Each constellation of magnets creates a super-structure pattern of supervised magnetic signatures. These signatures constitute an unambiguous magnetic environment with respect to the moving sensor carrier. The localization algorithm learns the unique patterns of the scattered magnets during training and detects them from the ongoing streaming of data during localization. Our work innovates regarding two essential features: first, instead of relying on active magnetic transmitters, we deploy passive permanent magnets that do not require a power supply. Second, we perform localization based on smartphone motion rather than on static positioning of the magnetometer. Therefore, we present a novel and unique dynamic indoor localization method combined with artificial intelligence (AI) techniques for post-processing. Experimental results have demonstrated localization accuracy of 95% with a resolution of less than 1m.
... probabilistic methods assume likelihoods of signal measurements at different RPs. Probabilistic algorithms such as maximum likelihood [18], expectation-maximization [12], and conditional random field [19] are often implemented in fingerprint-based indoor localization. Machine learning methods (e.g., random forest [20] and neural networks [21]) train a classification model to learn the signal-location relationship, and can achieve better localization performance at the expense of computational complexity. ...
Article
Transfer learning algorithms (TLAs) are often used to solve the distribution discrepancy issue in fingerprint-based indoor localization. However, existing TLAs cannot react well to real time changes in the environmental dynamics of the target space due to three remarkable shortcomings: a) redundant knowledge in source domain may lead to “negative transfer”; b) the required target domain samples to calculate the distributions are unrealistically feasible for real-time positioning; c) they cannot transfer knowledge efficiently across domains with heterogeneous feature spaces. In this paper, we propose TransLoc, a heterogeneous knowledge transfer framework for fingerprint-based indoor localization, which can perform knowledge transfer efficiently even with only one sample in the target domain. Specifically, we first refine the source domain according to the target domain by removing redundant knowledge in the source domain. Then, we derive a cross-domain mapping, which transfers the specific knowledge of one domain to another domain, to construct a homogeneous feature space. In this new feature space, the transfer weights are computed for training a classifier for target location prediction. To further train the framework efficiently, we combine the mapping and weights learning into a joint objective function and solve it by a three-step iterative optimization algorithm. Extensive simulation and real-world experimental results verify that TransLoc not only significantly outperforms state-of-the-art methods but is also very robust to changing environment.
... Compared to point-to-point matching, trajectory matching is more robust and has smaller matching error, but it is more complex and has poor real-time capability. Probabilistic graphical model-based map matching calculates the location by associating each location with a probability and then updating the probability using spatial constraints [7,13,20]. Probabilistic graphical mode-based approaches can achieve higher accuracy than point-to-point matching and trajectory matching, but its computational burden is heavier. ...
Article
Full-text available
Map-matching is a popular method that uses spatial information to improve the accuracy of positioning methods. The performance of map matching methods is closely related to spatial characteristics. Although several studies have demonstrated that certain map matching algorithms are affected by some spatial structures (eg, parallel paths), they focus on the analysis of single map matching method or few spatial structures. In this study, we explored how the most commonly-used four spatial characteristics (namely forks, open spaces, corners, and narrow corridors) affect three popular map matching methods, namely particle filtering (PF), hidden Markov model (HMM), and geometric methods. We first provide a theoretical analysis on how spatial characteristics affect the performance of map matching methods, and then evaluate these effects through experiments. We found that corners and narrow corridors are helpful in improving the positioning accuracy, while forks and open spaces often lead to a larger positioning error. We hope that our findings are helpful for future researchers in choosing proper map matching algorithms with considering the spatial characteristics.
... Wang et al. [8] presented UnLoc, established on landmarks matching. MapCraft, introduced by Xiao et al. [9], is rooted on the conditional random-fields method. The IODetector from Li et al. [10] used joint thresholds. ...
Article
Full-text available
Smartphones have become a popular tool for indoor localization and position estimation of users. Existing solutions mainly employ Wi-Fi, RFID, and magnetic sensing techniques to track movements in crowded venues. These are highly sensitive to magnetic clutters and depend on local ambient magnetic fields, which frequently degrades their performance. Also, these techniques often require pre-known mapping surveys of the area, or the presence of active beacons, which are not always available. We embed small-volume and large-moment magnets in pre-known locations and arrange them in specific geometric constellations that create magnetic superstructure patterns of supervised magnetic signatures. These signatures constitute an unambiguous magnetic environment with respect to the moving sensor carrier. The localization algorithm learns the unique patterns of the scattered magnets during training and detects them from the ongoing streaming of data during localization. Our contribution is twofold. First, we deploy passive permanent magnets that do not require a power supply, in contrast to active magnetic transmitters. Second, we perform localization based on smartphone motion rather than on static positioning of the magnetometer. In our previous study, we considered a single superstructure pattern. Here, we present an extended version of that algorithm for multi-superstructure localization, which covers a broader localization area of the user. Experimental results demonstrate localization accuracy of 95% with a mean localization error of less than 1m using artificial intelligence.
... Therefore, probabilistic algorithms record and store the RSSI distribution at each RP and use the probability distribution information for estimation, such as Horus [20]. More sophisticated probabilistic techniques have been recently investigated including Kernelized-based method [21], Kullback-Leibler (KL) divergence [22], Principal Component Analysis (PCA) [23], Conditional Random Fields (CRF) [24] and Bayesian Networks [25]. Comparably, probabilistic methods usually achieve higher localization accuracy as they exploit more accurate statistical representation of RSSI measurements. ...
Article
Full-text available
Received Signal Strength Indicator (RSSI) fingerprinting is known as the most concerned method for indoor localization as its high accuracy and low cost. Numerous RSSI based methods have shown their attractive performances, but the major drawback is the high dependency on the database. In this paper, we propose a multi-floor indoor localization method which includes floor detection and location estimation based on the radial basis function (RBF) network. To ensure the localization accuracy and stability, the network is constructed according to the probabilistic algorithm. Choosing Gaussian radial basis functions with appropriate widths, the network parameters can be initialized appropriately regardless of deficiency of RSSI data. By further conducting the supervised learning of RBF network, the network parameters will be effectively calibrated and updated, so as to achieve a better localization performance. In addition, a radio map quality evaluation criterion is proposed to conduct a comprehensive analysis and interpretation for the localization approach. Finally, experimental results of a publicly accessible dataset which includes multi-floors buildings verify that the performance of the proposed RBF network is superior to other commonly used methods.
... Wang et al. [8] presented UnLoc, established on landmarks matching. MapCraft, introduced by Xiao et al. [9], is rooted on the conditional random-fields method. The IODetector from Li et al. [10] used joint thresholds. ...
Preprint
Full-text available
Smartphones have become a popular tool for indoor localization and position estimation of users. Existing solutions mainly employ Wi-Fi, RFID, and magnetic sensing techniques to track movements in crowded venues. These are highly sensitive to magnetic clutters and depend on local ambient magnetic fields, which frequently degrades their performance. Also, these techniques often require pre-known mapping surveys of the area, or the presence of active beacons, which are not always available. We embed small-volume and large-moment magnets in pre-known locations and arrange them in specific geometric constellations that create magnetic superstructure patterns of supervised magnetic signatures. These signatures constitute an unambiguous magnetic environment with respect to the moving sensor carrier. The localization algorithm learns the unique patterns of the scattered magnets during training and detects them from the ongoing streaming of data during localization. Our contribution is twofold. First, we deploy passive permanent magnets that do not require a power supply, in contrast to active magnetic transmitters. Second, we perform localization based on smartphone motion rather than on static positioning of the magnetometer. In our previous study, we considered a single superstructure pattern. Here, we present an extended version of that algorithm for multi-superstructure localization, which covers a broader localization area of the user. Experimental results demonstrate localization accuracy of 95% with a mean localization error of less than 1m using artificial intelligence.
... Bojja et al. [33] extended the particle filter to three dimensions and combined it with collision detection techniques to navigate and localize vehicles in a parking garage. Other probabilistic approaches, such as the Kalman filter [10] and conditional random field (CRF) [34], are often used instead of a particle filter to reckon the appropriate localization. ...
Article
Full-text available
In recent years, using smartphones to determine pedestrian locations in indoor environments is an extensively promising technique for improving context-aware applications. However, the applicability and accuracy of the conventional approaches are still limited due to infrastructure-dependence, and there is seldom consideration of the semantic information inherently existing in maps. In this paper, a semantically-constrained low-complexity sensor fusion approach is proposed for the estimation of the user trajectory within the framework of the smartphone-based indoor pedestrian localization, which takes into account the semantic information of indoor space and its compatibility with user motions. The user trajectory is established by pedestrian dead reckoning (PDR) from the mobile inertial sensors, in which the proposed semantic augmented route network graph with adaptive edge length is utilized to provide semantic constraint for the trajectory calibration using a particle filter algorithm. The merit of the proposed method is that it not only exploits the knowledge of the indoor space topology, but also exhausts the rich semantic information and the user motion in a specific indoor space for PDR accumulation error elimination, and can extend the applicability for diverse pedestrian step length modes. Two experiments are conducted in the real indoor environment to verify of the proposed approach. The results confirmed that the proposed method can achieve highly acceptable pedestrian localization results using only the accelerometer and gyroscope embedded in the phones, while maintaining an enhanced accuracy of 1.23 m, with the indoor semantic information attached to each pedestrian’s motion.
... The first approach is mainly focused on filtering the sensor values. Various researches have been done so far in [41][42][43][44][45][46]. Most studies have improved positioning accuracy by filtering the values of access point (AP) signal or smartphone sensors (acceleration, gyroscope, etc.) used in the pedestrian dead reckoning (PDR) method. ...
Article
Full-text available
With the development of indoor positioning methods, such as Wi-Fi positioning, geomagnetic sensor positioning, Ultra-Wideband positioning, and pedestrian dead reckoning, the area of location-based services (LBS) is expanding from outdoor to indoor spaces. LBS refers to the geographic location information of moving objects to provide the desired services. Most Wi-Fi-based indoor positioning methods provide two-dimensional (2D) or three-dimensional (3D) coordinates in 1–5 m of accuracy on average approximately. However, many applications of indoor LBS are targeted to specific spaces such as rooms, corridors, stairs, etc. Thus, they require determining a service space from a coordinate in indoor spaces. In this paper, we propose a map matching method to assign an indoor position to a unit space a subdivision of an indoor space, called USMM (Unit Space Map Matching). Map matching is a commonly used localization improvement method that utilizes spatial constraints. We consider the topological information between unit spaces and moving objects’ probabilistic properties, compared to existing room-level mappings based on sensor signals, especially received signal strength-based fingerprinting. The proposed method has the advantage of calculating the probability even if there is only one input trajectory. Last, we analyze the accuracy and performance of the proposed USMM methods by extensive experiments in real and synthetic environments. The experimental results show that our methods bring a significant improvement when the accuracy level of indoor positioning is low. In experiments, the room-level location accuracy improves by almost 30% and 23% with real and synthetic data, respectively. We conclude that USMM methods are helpful to correct valid room-level locations from given positioning locations.
... In [2] additional hardware is needed for construction the map of the floor, a robot equipped with a laser rangefinder and fiber optic gyroscope is used for providing an automated procedure for collecting Wi-Fi calibration data to generate the robot map. MapCraft [3] uses map matching algorithm and undirected graphical model, known as linear chain conditional random fields for dealing with floorplan which increases the complexity of the implementation. In MapCraft, the floorplan is taken as input to the system then a number of complex information is implemented to define the floorplan to the system and these steps are repeated for every new floorplan. ...
Article
Full-text available
Indoor maps are highly essential for indoor positioning and location-based services. People existing in big building such as mall, airport and hospitals leads to rapid growth in location based applications. Todays, typical indoor positioning systems employ a training/positioning model using Wi-Fi fingerprints. While these approaches have practical results in terms of coverage, they require an indoor map, which is rarely available to the user and involves significant training costs. In this paper, we present a virtual map, generated automatically inside user smartphone, which lets the user watching her/himself moving forward, backward, left or right from well-known landmarks on the floor map. This map is generated in the form of rectangle of grid points with no prior information of the actual indoor map; therefore, it is suitable for using it at any building. An indoor positioning system based on Smartphone sensor (Accelerometer and Magnetometer) incorporated with existing APs in the building is used to test the suggested virtual map.
... Several sophisticated probabilistic techniques, such as principal component localization, 21 conditional random field 22 and the Bayesian network, have been studied. Madigan et al. 23 introduced a 2D Bayesian system based on the probabilistic approach in which the number of samples is generated from the posterior distribution using Gibbs sampling to predict the users' location based on the maximum posteriori. ...
Article
Full-text available
Access points in wireless local area networks are deployed in many indoor environments. Device-free wireless localization systems based on available received signal strength indicators have gained considerable attention recently because they can localize the people using commercial off-the-shelf equipment. Majority of localization algorithms consider two-dimensional models that cause low positioning accuracy. Although three-dimensional localization models are available, they possess high computational and localization errors, given their use of numerous reference points. In this work, we propose a three-dimensional indoor localization system based on a Bayesian graphical model. The proposed model has been tested through experiments based on fingerprinting technique which collects received signal strength indicators from each access point in an offline training phase and then estimates the user location in an online localization phase. Results indicate that the proposed model achieves a high localization accuracy of more than 25% using reference points fewer than that of benchmarked algorithms.
... Another type of method is called the probability method, which considers the position fingerprint's randomness and maps a fingerprint into the probability density of the position. The usual methods are naive Bayes [24], probability kernel regression [25], conditional random field [26], and so on. ...
Article
Full-text available
Wireless fingerprinting localization (FL) systems identify locations by building radio fingerprint maps, aiming to provide satisfactory location solutions for the complex environment. However, the radio map is easy to change, and the cost of building a new one is high. One research focus is to transfer knowledge from the old radio maps to a new one. Feature-based transfer learning methods help by mapping the source fingerprint and the target fingerprint to a common hidden domain, then minimize the maximum mean difference (MMD) distance between the empirical distributions in the latent domain. In this paper, the optimal transport (OT)-based transfer learning is adopted to directly map the fingerprint from the source domain to the target domain by minimizing the Wasserstein distance so that the data distribution of the two domains can be better matched and the positioning performance in the target domain is improved. Two channel-models are used to simulate the transfer scenarios, and the public measured data test further verifies that the transfer learning based on OT has better accuracy and performance when the radio map changes in FL, indicating the importance of the method in this field.
... Similarly, WiFi-SLAM and Zee build on particle filters emphasising their importance for random system initialisation [3], while Kalman filters are used to integrate inertial sensing modalities [5]. Other engineered ap-proaches, such as UnLoc, combine sensing modalities based on empirical observations of how some locations are unique across one or more sensors [6], MapCraft uses conditional random fields [23] and LiFS uses graph constraints to map and position estimations on the trajectory [4]. Similarly, WILL builds a connected graph to estimate location at room level [24]. ...
Article
Full-text available
Many engineered approaches have been proposed over the years for solving the hard problem of performing indoor localization using smartphone sensors. However, specialising these solutions for difficult edge cases remains challenging. Here we propose an end-to-end hybrid multimodal deep neural network localization system, MM-Loc, relying on zero hand-engineered features, but learning automatically from data instead. This is achieved by using modality-specific neural networks to extract preliminary features from each sensing modality, which are then combined by cross-modality neural structures. We show that our choice of modality-specific neural architectures can estimate the location independently. But for better accuracy, a multimodal neural network that fuses the features of early modality-specific representations is a better proposition. Our proposed MM-Loc system is tested on cross-modality samples characterised by different sampling rate and data representation (inertial sensors, magnetic and WiFi signals), outperforming traditional approaches for location estimation. MM-Loc elegantly trains directly from data unlike conventional indoor positioning systems, which rely on human intuition.
... Similarly, WiFi-SLAM and Zee build on particle filters emphasising their importance for random system initialisation [5], while Kalman filters are used to integrate inertial sensing modalities [6]. Other engineered approaches, such as UnLoc, combining sensing modalities based on empirical observations of how some locations are unique across one or more sensors [2], MapCraft uses conditional random fields [16], LiFS uses graph constraints to map and position estimations on the trajectory [4]. Similarly, WILL builds a connected graph to estimate location at room level [17]. ...
Conference Paper
Indoor positioning systems have been explored for decades to facilitate universal location-based services. However, complex environment conditions and sensing imperfections continue to be limiting factors to their large scale adoption. Rather than customising more ingenious solutions to handle corner cases in complex environments, we believe that a more efficient solution is to learn entirely from data with minimal engineering effort. We develop neural network based solutions for two positioning approaches, modelling Dead Reckoning with recurrent neural networks and WiFi Fingerprinting with deep neural networks. We propose a multimodal deep neural network architecture (MM-Loc) that bridges the features extracted by the modality-specific complements (sensor based and WiFi based) to join the two perspectives. We observe that this multimodal approach is better than single-modality models, and elegantly trains directly from raw data with minimal intervention.
... • Conditional Random Field (CRF) is based on a stateof-the-art map-matching system [28], which computes a reach-ability graph from a floorplan, uses inertial navigation results to transition between graph nodes, and uses Viterbi algorithm to backtrack and determine location. We modified the system in a few ways to better adapt to our Table 3. Ablation study. ...
Preprint
Full-text available
This paper proposes the inertial localization problem, the task of estimating the absolute location from a sequence of inertial sensor measurements. This is an exciting and unexplored area of indoor localization research, where we present a rich dataset with 53 hours of inertial sensor data and the associated ground truth locations. We developed a solution, dubbed neural inertial localization (NILoc) which 1) uses a neural inertial navigation technique to turn inertial sensor history to a sequence of velocity vectors; then 2) employs a transformer-based neural architecture to find the device location from the sequence of velocities. We only use an IMU sensor, which is energy efficient and privacy preserving compared to WiFi, cameras, and other data sources. Our approach is significantly faster and achieves competitive results even compared with state-of-the-art methods that require a floorplan and run 20 to 30 times slower. We share our code, model and data at https://sachini.github.io/niloc.
... Some works leverage this deployment information to improve localization accuracy. For example, MapCraft [19] uses the signal from the strongest access point as a landmark and constraints the location estimation to a subset of access points. HALLWAY [45], on the other hand, uses a subset of access points, in which their signal strengths are in ascending order, to first classify a coarse-grain area prior to estimating a fine-grain location. ...
... Some works leverage this deployment information to improve localization accuracy. For example, MapCraft [21] uses the signal from the strongest access point as a landmark and constrains the location estimation to a subset of access points. HALLWAY [47], on the other hand, uses a subset of access points, in which their signal strengths are in ascending order, to first classify a coarse-grain area prior to estimating a fine-grain location. ...
Preprint
Full-text available
This paper presents a nonlinear location estimation to infer the position of a user holding a smartphone. We consider a large location with $M$ number of grid points, each grid point is labeled with a unique fingerprint consisting of the received signal strength (RSS) values measured from $N$ number of Bluetooth Low Energy (BLE) beacons. Given the fingerprint observed by the smartphone, the user's current location can be estimated by finding the top-k similar fingerprints from the list of fingerprints registered in the database. Besides the environmental factors, the dynamicity in holding the smartphone is another source to the variation in fingerprint measurements, yet there are not many studies addressing the fingerprint variability due to dynamic smartphone positions held by human hands during online detection. To this end, we propose a nonlinear location estimation using the kernel method. Specifically, our proposed method comprises of two steps: 1) a beacon selection strategy to select a subset of beacons that is insensitive to the subtle change of holding positions, and 2) a kernel method to compute the similarity between this subset of observed signals and all the fingerprints registered in the database. The experimental results based on large-scale data collected in a complex building indicate a substantial performance gain of our proposed approach in comparison to state-of-the-art methods. The dataset consisting of the signal information collected from the beacons is available online.
... Some other more advanced deterministic algorithms such as Support Vector Machine (SVM) [74,66] and Linear Discriminant Analysis (LDA) [75] show better localization accuracy with higher computational cost. Other probabilistic algorithms such as Bayesian network [76,77], expectation maximization [78], Kullback-Leibler Divergence (KLD) [79,33], Gaussian process [80] and conditional random field [81] can also achieve high localization accuracy through probabilistic inference. ...
Thesis
With expeditious development of wireless communications, Location Fingerprinting (LF) has nurtured considerable indoor location based services in the field of Internet of Things. In this thesis, we first proposed EntLoc system, which adopts Autoregressive (AR) modeling entropy of the Channel State Information (CSI) amplitude as location fingerprint. It shares the structural simplicity of the Received Signal Strength (RSS) while reserving the most location-specific statistical channel information. Moreover, an upgraded AngLoc system is further designed, whose additional angle of arrival (AoA) fingerprint can be accurately retrieved from CSI phase through an enhanced subspace based algorithm, which serves to further eliminate the error-prone Reference Point (RP) candidates. In the LF online phase, by exploiting both CSI amplitude and phase information, a novel bivariate kernel regression scheme is proposed to precisely infer the target’s location. Results from extensive indoor experiments validate the superior localization performance of our proposed system over previous approaches.
... Recently, with the research of conditional random field model algorithm, it has made significant progress in map matching. In 2014, Xiao and Zhuoling from the University of Oxford applied the conditional random field (CRF) model to map matching to improve the inertial localization [24]. Study [25] has proved that compared with other algorithms, the CRF model can capture a variety of constraints relationships between observations, with more universal applicability and higher accuracy. ...
Article
Full-text available
High-precision indoor localization plays a vital role in various places. In recent years, visual inertial odometry (VIO) system has achieved outstanding progress in the field of indoor localization. However, it is easily affected by poor lighting and featureless environments. For this problem, we propose an indoor localization algorithm based on VIO system and three-dimensional (3D) map matching. The 3D map matching is to add height matching on the basis of previous two-dimensional (2D) matching so that the algorithm has more universal applicability. Firstly, the conditional random field model is established. Secondly, an indoor three-dimensional digital map is used as a priori information. Thirdly, the pose and position information output by the VIO system are used as the observation information of the conditional random field (CRF). Finally, the optimal states sequence is obtained and employed as the feedback information to correct the trajectory of VIO system. Experimental results show that our algorithm can effectively improve the positioning accuracy of VIO system in the indoor area of poor lighting and featureless.
Article
Metric is not only a function to mesure the distance between data points, but also a main tool to evaluate the error of data analysis, so it is one of the important factors to be considered by expert systems and intelligent systems. A general metric function is difficult to adapt to all application scenarios. Thanks to the development of big data and machine learning technology, metric learning can be used to obtain an optimal metric for a specific application scenario. Considering the fingerprinting localization (FL) as an intelligent system for processing radio signals, this paper proposes two novel novel location fingerprint (LF) metric learning algorithms to improve the accuracy and adaptability of the system. The two proposed algorithms are named LMNN-LF and NCA-LF respectively, and they are based on two famous metric learning algorithms, LMNN and NCA. Considering the distribution characteristics of position fingerprints, we modified the original cost function in both schemes. In order to accommodate the low resolution of the fingerprint, an error radius is defined in our method. The distance relation of the LFs in high dimensional space is better described by the learned metric function, which improves the accuracy of the FL system compared with the general metric function or metric learning method. Moreover, the proposed methods can also effectively extract the features or reduce the dimension of LFs, improving the accuracy of other feature-based localization algorithms. Experiments on different data sets show that the metric obtained by the proposed method and several existing methods performs good in kNN localization, and performs good in LF feature extraction and dimensionality reduction.
Article
Indoor localization has attracted more and more attention because of its importance in many applications. One of the most popular techniques for indoor localization is the received signal strength indicator (RSSI) based fingerprinting approach. Since RSSI values are very complicated and noisy, conventional machine learning algorithms often suffer from limited performance. Recently developed deep learning algorithms have been shown to be powerful for the analysis of complex data. In this paper, we propose a local feature-based deep long short-term memory (LF-DLSTM) approach for WiFi fingerprinting indoor localization. The local feature extractor attempts to reduce the noise effect and extract robust local features. The DLSTM network is able to encode temporal dependencies and learn high-level representations for the extracted sequential local features. Real experiments have been conducted in two different environments, i.e., a research lab and an office. We also compare the proposed approach with some state-of-the-art methods for indoor localization. The results show that the proposed approach achieves the best localization performance with mean localization errors of 1.48 and 1.75 m under the research lab and office environments, respectively. The improvements of our proposed approach over the state-of-the-art methods range from 18.98% to 53.46%.
Article
Location-based services have become extremely popular. As a result, indoor localization has received a lot of attention from both industry and academia. Despite the plethora of existing approaches, no method can achieve high accuracy without major, expensive changes to existing systems. For example, to achieve good accuracy, fingerprinting methods require many APs within range of a client, time-of-arrival methods require tight synchronization, client hardware changes and line-of-sight, and direction-of-arrival methods require expensive antenna front ends and line-of-sight. In this paper, we take advantage of inexpensive, off-the-shelf switched-beam antennas (SBAs) to increase the diversity of measurements used for fingerprint-based localization. We show using experiments that a single AP equipped with n SBAs may infer equally rich localization information as nAPs equipped with n omnidirectional antennas each, as long as the SBAs are properly configured. We then establish via extensive experiments that a single packet reception from a commodity client at a single AP equipped with a handful of SBAs (e.g., eight SBAs costing a couple of dollars more than omni antennas) achieves localization accuracy in the order of half a meter with or without line-of-sight, in any indoor environment, with zero airtime overhead and zero client support.
Chapter
With the ongoing diffusion of mobile computing and context-aware applications, knowledge of the current location of an individual can be leveraged in a number of different domains, from personal diaries and fitness-related applications to human behavior analysis and targeted advertising. This chapter presents a review of past research works describing techniques for utilizing smartphone sensors to identify the environment where a smartphone user is located. The review focuses on studies where user location can be computed autonomously and continuously by a smartphone without the need for an active involvement of the user, and where issues such as power consumption and dependence of sensor readings from the on-body position of the phone are addressed.
Article
Indoor positioning complements the mature outdoor positioning technology, Global Navigation Satellite System (GNSS), by achieving real-time positioning in any environment under a blockage of GNSS signals. In the construction management field, this paper demonstrates that indoor positioning enables five significant applications that considerably enhance work efficiency and safety on construction sites. Without a perfect indoor positioning system and under complicated site environment, developing a suitable on-site indoor positioning system is challenging and essentially user-oriented and environment-specific. This paper analyses the challenges to implement on-site indoor positioning systems, and proposes indoor positioning performance metrics, namely APP-CAT, for evaluating indoor positioning systems. The fundamental indoor positioning principles are first discussed and evaluated. Subsequently, the top 10 indoor positioning technologies, selected by their performance in APP-CAT and their popularity, are thoroughly compared. The promising trends of indoor positioning development, e.g., indoor positioning hybridization, game theory positioning, and integration with BIM models, are highlighted.
Article
map matching has played a crucial role in technologies related to indoor positioning. Conventional map matching algorithms based on particle filter (PF) have some limitations, such as the limited use of map information, poor generalization and low precision. To solve these problems, we propose an adaptable particle filter network (AdaPFnet), a novel map matching technique that integrates particle filter algorithm into a neural network. AdaPFnet uses local views of particles to represent particles so that the map information about location can be learned sufficiently through a neural network. To demonstrate the performance of the model, it has conducted extensive experiments using 1540 real-world data. The results show that AdaPFnet outperforms PF by up to 21% and remains a strong adaptability for different environments.
Article
Smart scheduling of domestic appliances based on user demand is an important aspect of any home energy management system. Such a demand driven operation can be performed only by accurately detecting the presence and the location of users in their residence. IoT and other enabling technologies can transform a smart home into an energy aware entity by judicious utilization of the available data. This paper presents a survey of various occupancy detection and localization schemes and evaluates them on the basis of various factors that decide their suitability for home energy management system. Wireless technologies have been historically employed for detecting unknown targets in indoor environments. The pervasiveness of WiFi access points in urban buildings render it as a preferred candidate for occupancy detection and localization. Design of device-free sensing techniques, using commercially available WiFi routers can offer an affordable solution with minimum disturbance to the inhabitants. Detection accuracy maybe enhanced by combining wireless techniques with smartphone inertial sensors and other passive detection schemes that utilize smart energy meters or environmental sensors. In contrast to commercial buildings, several challenges must be addressed for occupancy detection in home energy management system. The impact of these design challenges are investigated and feasible solutions are suggested.
Article
Natural navigation simply refers to free navigation without the necessity of tapes, magnets, reflectors, or even wires. Many autonomous vehicles possess this as world maps are readily available and provide a perfect basis for machine learning solutions. However, this is not so much the case for indoor applications. Here, paths are often dynamic and more constrained; therefore, requiring the continuous identification, mapping and localization of the surrounding area. This work focuses on developing an indoor natural navigation system; the localization is achieved with a fusion of the wheel’s odometry to the on-board Inertial Measurement Unit (IMU i.e., a combination of relative localization and absolute localization) using Unscented Kalman Filter (UKF) as system’s encoder’s accumulation of errors is desired to be nullified while employing a PID control in correcting reference state errors. The map is simultaneously constructed using laws of geometry based on static points obtained from a Lidar, subsequently converted to an occupancy grid layout for effective path planning. In operation, tangency is applied in the avoidance of dynamic obstacles. The simulation results obtained in this study confirms the possibility of a simple, educational, indoor navigation system approach easily integrable by other mobile robots of the differential drive model.
Article
Full-text available
Inertial odometry is a typical localization method that is widely and easily accessible in many devices. Pedestrian positioning can benefit from this approach based on inertial measurement unit (IMU) values embedded in smartphones. Fitting the inertial odometry outputs, namely step length and step heading of a human for instance, with spatial information is an ubiquitous way to correct for the cumulative noises. This so-called map-matching process can be achieved in several ways. In this paper, a novel real-time map-matching approach was developed, using a backtracking particle filter that benefits from the implemented geospatial analysis, which reduces the complexity of spatial queries and provides flexibility in the use of different kinds of spatial constraints. The goal was to generalize the algorithm to permit the use of any kind of odometry data calculated by different sensors and approaches as the input. Further research, development, and comparisons have been done by the easy implementation of different spatial constraints and use cases due to the modular structure. Additionally, a simple map-based optimization using transition areas between floors has been developed. The developed algorithm could achieve accuracies of up to 3 m at approximately the 90th percentile for two different experiments in a complex building structure.
Article
5G technologies promise improved features for positioning accuracy for smart city use cases in indoor and outdoor environments. This paper presents an overview of the existing positioning technologies and their accuracy, as well as the potential role of 5G to enhance the development of enabled use cases in smart cities. Therefore, based on the specifications of each use case, the author analyzes and discusses the impacts and roles of 5G features on the emerging positioning methods. Furthermore, solutions were suggested to improve the development of 5G-based use cases in smart cities.
Thesis
Industry 4.0 is driving change for new forms of production and real-time optimization in factories, which benefit from the Industrial Internet of Things (IoT) capabilities to locate industrial vehicles for monitoring, improving safety, and operations. Most industrial environments have a Wi-Fi infrastructure that can be exploited to locate people, assets, or vehicles, providing an opportunity for enhancing productivity and interconnectivity. Radio maps are important for Wi-Fi-based Indoor Position Systems (IPSs) since they represent the radio environment and are used to estimate a position. Radio maps comprise a set of Wi- Fi samples collected at known positions, and degrade over time due to several aspects, e.g., propagation effects, addition/removal of Access Points (APs), among others, hence they should be periodically updated to maintain the IPS performance. The process to build and maintain radio maps is usually time-consuming and demanding in terms of human resources, thus being challenging to perform. Vehicles, commonly present in industrial environments, can be explored to help build and maintain radio maps, as long as it is possible to locate and track them. The main objective of this thesis is to develop an IPS for industrial vehicles with self-healing radio maps (capable of keeping radio maps up to date). Vehicles are tracked using sensor fusion of Wi-Fi with motion sensors, which allows to annotate new Wi-Fi samples to build the self-healing radio maps. Two sensor fusion approaches based on Loose Coupling and Tight Coupling are proposed to track vehicles. The Tight Coupling approach includes a reliability metric to determine when Wi-Fi samples should be annotated. As a result, this solution does not depend on any calibration or human effort to build and maintain the radio map. Results obtained in real-world experiments suggest that this solution has potential for IoT and Industry 4.0, especially in location services, but also in monitoring and analytics, supporting autonomous navigation, and interconnectivity between devices.
Article
Localisation is an important part of many applications. Our motivating scenarios are short-term construction work and emergency rescue. These scenarios also require rapid setup and robustness to environmental conditions additional to localisation accuracy. These requirements preclude the use of many traditional high-performance methods, e.g. vision-based, laser-based, Ultra-wide band (UWB) and Global Positioning System (GPS)-based localisation systems. To overcome these challenges, we introduce iMag+, an accurate and rapidly deployable inertial magneto-inductive (MI) mapping and localisation system, which only requires monitored workers to carry a single MI transmitter and an inertial measurement unit in order to localise themselves with minimal setup effort. However, one major challenge is to use distorted and ambiguous MI location estimates for localisation. To solve this challenge, we propose a novel method to use MI devices for sensing environmental distortions for accurate closing inertial loops. We also suggest a robust and efficient first quadrant estimator to sanitise the ambiguous MI estimates. By applying robust simultaneous localisation and mapping (SLAM), our proposed localisation method achieves excellent tracking accuracy and can improve performance significantly compared with only using a Magneto-inductive device or inertial measurement unit (IMU) for localisation.
Chapter
Human mobility estimation provides sufficient information for urban planning and transportation management, which plays a significant role in improving urban mobility, accessibility, and quality of residents’ life. Call Detail Record, which is a kind of mobile phone data, could be utilized to estimate human mobility, but face with the problems that it is hard to extract accurate position from coarse indicated range. A method is introduced to demonstrate the standard process to address this challenge, which could filter raw CDR, process it into trip segments based on the moving mode and combine map-matching with interpolation for different moving modes to estimate the real movement. A field research CDR dataset collected in Hongkong, China and a simulated CDR dataset based on a real GPS dataset in Tokyo, Japan will be utilized to validate the estimation result in different CDR sampling conditions to show the accuracy of the method and influence of the sampling conditions.
Article
Unsupervised localization based on received signal strength and inertial measurement unit (RSS+IMU) sequences is an important branch of indoor localization community, among which transitional model to predict motion from signal change (TMM) is a promising method that does not require much prior knowledge, e.g., floor maps. However, there are still many challenging problems existing in TMM. Among them, the computation burden of the model is awfully heavy, and its localization error is also a painful point. In order to solve the above challenges, we propose a novel transition model, called Enhanced TMM (ETMM). 1) Trajectory Data Enhancement is proposed to enrich the diversity of trajectory data, which improves the robustness of the transition model by allowing the model to learn more comprehensive and detailed information from the environment. 2) The computation burden has been significantly reduced by using Effective RSS Preprocessing which reduces the data dimension and the solution domain. 3) In order to increase the robustness and localization accuracy of the model, we propose Direction Matching (DM) constraint to enhance the mapping relationship between the consecutive RSS signals and the one-step motion. Experiments show that ETMM has better performance compared with the state-of-the-art method in terms of localization accuracy, computation cost, and robustness.
Article
The fingerprinting-based positioning has great potential for indoor location estimation where GPS signals are mostly blocked. However, fingerprinting-based methods need a calibration step for establishing a fingerprint map. The site survey process should be performed to record fingerprints which is a labor-intensive task essentially in large buildings. In this paper, we address the pedestrian trajectory reconstruction problem for fingerprint map creation. Where the goal is to predict and refine users’ trajectories obtained from smartphone sensor measurements. Our proposed spatial–temporal matching mechanism consists of three stages. First, the initial trajectory is calculated using the PDR algorithm. Then, landmarks error is eliminated using the proposed forward/backward error correction (FEC/BEC) algorithm. Afterward, the proposed dynamic time warping-based path-matching (DTW-PM) method applies to handle map-related errors. The evaluation results show positioning accuracy improves up to 53.69%. Finally, a traditional KNN algorithm is performed to evaluate the positioning efficiency over generated radio map, which validates the quality of the obtained radio map.
Article
Although vehicle location-based services are prevalent outdoors, we are back into darkness in many GPS blocked environments such as tunnels, indoor parking garages, and multilevel flyovers. Existing smartphone-based solutions usually adopt inertial dead-reckoning to infer the trajectory, but low-quality inertial sensors in phones are plagued by heavy noises, causing unbounded localization errors through double integrations for movements. In this paper, we propose VeTorch, a smartphone inertial odometry that devises an inertial sequence learning framework to track vehicles in real time when GPS signal is not available. Specifically, we transform the inertial dynamics from the phone to the vehicle regardless of arbitrary phone’s placement in car, and explore a temporal convolutional network to learn vehicle’s moving dependencies directly from the inertial data. To tackle the heterogeneous smartphone properties and driving habits, we propose a federated learning based active model training mechanism to produce customized models for individual smartphones, without incurring user privacy issues. We implement a highly efficient prototype and conduct extensive experiments on two large-scale real-world traffic datasets collected by a modern ride-hailing platform. Our results outperform the state-of-the-art vehicular inertial dead-reckoning solutions on both accuracy and efficiency.
Article
Current localization schemes on mobile devices are experiencing great diversity that is mainly shown in two aspects: the large number of available localization schemes and their diverse performance. This paper presents UniLoc , a unified framework that gains improved performance from multiple localization schemes by exploiting their diversity. UniLoc predicts the localization error of each scheme online based on an error model and real-time context. It further combines the results of all available schemes based on the error prediction results and an ensemble learning algorithm. The combined result is more accurate than any individual schemes. With the flexible design of error modeling and ensemble learning, UniLoc can easily integrate a new localization scheme. The energy consumption of UniLoc is low, since its computation, including both error prediction and ensemble learning, only involves simple linear calculation. Our experience with extensive experiments tells that such easy aggregation incurs little overhead in integrating and training a localization scheme, but gains substantially from the scheme diversity. UniLoc outperforms individual localization schemes by 1.6× in a variety of environments, including $>89\%$ new places where we did not train the error models.
Conference Paper
Full-text available
This paper addresses reliable and accurate indoor localization using inertial sensors commonly found on commodity smartphones. We believe indoor positioning is an important primitive that can enable many ubiquitous computing applications. To tackle the challenges of drifting in estimation, sensitivity to phone position, as well as variability in user walking profiles, we have developed algorithms for reliable detection of steps and heading directions, and accurate estimation and personalization of step length. We've built an end-to-end localization system integrating these modules and an indoor floor map, without the need for infrastructure assistance. We demonstrated for the first time a meter-level indoor positioning system that is infrastructure free, phone position independent, user adaptive, and easy to deploy. We have conducted extensive experiments on users with smartphone devices, with over 50 subjects walking over an aggregate distance of over 40 kilometers. Evaluation results showed our system can achieve a mean accuracy of 1.5m for the in-hand case and 2m for the in-pocket case in a 31m×15m testing area.
Article
Full-text available
We propose UnLoc, an unsupervised indoor localization scheme that bypasses the need for war-driving. Our key observation is that certain locations in an indoor environment present identifiable signatures on one or more sensing dimensions. An elevator, for instance, imposes a distinct pattern on a smartphone's accelerometer; a corridor-corner may overhear a unique set of WiFi access points; a specific spot may experience an unusual magnetic fluctuation. We hypothesize that these kind of signatures naturally exist in the environment, and can be envisioned as internal landmarks of a building. Mobile devices that "sense" these landmarks can recalibrate their locations, while dead-reckoning schemes can track them between landmarks. Results from 3 different indoor settings, including a shopping mall, demonstrate median location errors of 1:69m. War-driving is not necessary, neither are floorplans the system simultaneously computes the locations of users and landmarks, in a manner that they converge reasonably quickly. We believe this is an unconventional approach to indoor localization, holding promise for real-world deployment.
Article
Full-text available
The major challenge for accurate fingerprint-based indoor localization is the design of robust and discriminative wireless signatures. Even though WiFi RSSI signatures are widely available indoors, they vary significantly over time and are susceptible to human presence, multipath, and fading due to the high operating frequency. To overcome these limitations, we propose to use FM broadcast radio signals for robust indoor fingerprinting. Because of the lower frequency, FM signals are less susceptible to human presence, multipath and fading, they exhibit exceptional indoor penetration, and according to our experimental study they vary less over time when compared to WiFi signals. In this work, we demonstrate through a detailed experimental study in 3 different buildings across the US, that FM radio signal RSSI values can be used to achieve room-level indoor localization with similar or better accuracy to the one achieved by WiFi signals. Furthermore, we propose to use additional signal quality indicators at the physical layer (i.e., SNR, multipath etc.) to augment the wireless signature, and show that localization accuracy can be further improved by more than 5%. More importantly, we experimentally demonstrate that the localization errors of FM andWiFi signals are independent. When FM and WiFi signals are combined to generate wireless fingerprints, the localization accuracy increases as much as 83% (when accounting for wireless signal temporal variations) compared to when WiFi RSSI only is used as a signature.
Article
Full-text available
Microelectromechanical Systems (MEMS) technology is playing a key role in the design of the new generation of smartphones. Thanks to their reduced size, reduced power consumption, MEMS sensors can be embedded in above mobile devices for increasing their functionalities. However, MEMS cannot allow accurate autonomous location without external updates, e.g., from GPS signals, since their signals are degraded by various errors. When these sensors are fixed on the user's foot, the stance phases of the foot can easily be determined and periodic Zero velocity UPdaTes (ZUPTs) are performed to bound the position error. When the sensor is in the hand, the situation becomes much more complex. First of all, the hand motion can be decoupled from the general motion of the user. Second, the characteristics of the inertial signals can differ depending on the carrying modes. Therefore, algorithms for characterizing the gait cycle of a pedestrian using a handheld device have been developed. A classifier able to detect motion modes typical for mobile phone users has been designed and implemented. According to the detected motion mode, adaptive step detection algorithms are applied. Success of the step detection process is found to be higher than 97% in all motion modes.
Article
Full-text available
In this paper a novel step length model using a handheld Micro Electrical Mechanical System (MEMS) is presented. It combines the user's step frequency and height with a set of three parameters for estimating step length. The model has been developed and trained using 12 different subjects: six men and six women. For reliable estimation of the step frequency with a handheld device, the frequency content of the handheld sensor's signal is extracted by applying the Short Time Fourier Transform (STFT) independently from the step detection process. The relationship between step and hand frequencies is analyzed for different hand's motions and sensor carrying modes. For this purpose, the frequency content of synchronized signals collected with two sensors placed in the hand and on the foot of a pedestrian has been extracted. Performance of the proposed step length model is assessed with several field tests involving 10 test subjects different from the above 12. The percentages of error over the travelled distance using universal parameters and a set of parameters calibrated for each subject are compared. The fitted solutions show an error between 2.5 and 5% of the travelled distance, which is comparable with that achieved by models proposed in the literature for body fixed sensors only.
Conference Paper
Full-text available
Human localization is a very valuable information for smart environments. state-of-the-art local positioning systems (LPS) require a complex sensor-network infrastructure to locate with enough accuracy and coverage. Alternatively, inertial measuring units (IMU) can be used to estimate the movement of a person, by detecting steps, estimating stride lengths and the directions of motion; a methodology that is called pedestrian dead-reckoning (PDR). In this paper, we use low-performance microelectromechanical (MEMS) inertial sensors attached to the foot of a person. This sensor has triaxial orthogonal accelerometers, gyroscopes and magnetometers. We describe, implement and compare several of the most relevant algorithms for step detection, stride length, heading and position estimation. The challenge using MEMS is to provide location estimations with enough accuracy and a limited drift. Several tests were conducted outdoors and indoors, and we found that the stride length estimation errors were about 1%. The positioning errors were almost always below 5% of the total travelled distance. The main source of positioning errors are the absolute orientation estimation.
Conference Paper
Full-text available
An algorithm for pedestrian navigation in indoor and urban canyon environments is presented. It considers platforms with low processing power and low-cost sensors. A combination of Wi-Fi positioning and dead reckoning, based on a Hidden Markov Model, is used. The positions of the Wi-Fi fingerprints in the database are used as hidden states. Dead reckoning is taken for state transition and a database correlation of the Wi-Fi signal strength measurements is performed in the measurement update. The dead reckoning consists of an accelerometer driven step length estimation and a magnetic field based heading calculation. Simulations and tests demonstrate that in this way ambiguities common in Wi-Fi positioning can be solved and outages can be bridged. Therefore, higher accuracy and robustness can be achieved.
Conference Paper
Full-text available
We present the design and implementation of the Horus WLAN location determination system. The design of the Horus system aims at satisfying two goals: high accuracy and low computational requirements. The Horus system identifies different causes for the wireless channel variations and addresses them to achieve its high accuracy. It uses location-clustering techniques to reduce the computational requirements of the algorithm. The lightweight Horus algorithm helps in supporting a larger number of users by running the algorithm at the clients.We discuss the different components of the Horus system and its implementation under two different operating systems and evaluate the performance of the Horus system on two testbeds. Our results show that the Horus system achieves its goal. It has an error of less than 0.6 meter on the average and its computational requirements are more than an order of magnitude better than other WLAN location determination systems. Moreover, the techniques developed in the context of the Horus system are general and can be applied to other WLAN location determination systems to enhance their accuracy. We also report lessons learned from experimenting with the Horus system and provide directions for future work.
Conference Paper
Full-text available
Traffic delays and congestion are a major source of ineffi- ciency, wasted fuel, and commuter frustration. Measuring and localizing these delays, and routing users around them, is an important step towards reducing the time people spend stuck in traffic. As others have noted, the proliferation of commod- ity smartphones that can provide location estimates using a variety of sensors—GPS, WiFi, and/or cellular triangulation— opens up the attractive possibility of using position samples from drivers' phones to monitor traffic delays at a fine spatio- temporal granularity. This paper presents VTrack, a system for travel time estimation using this sensor data that addresses two key challenges: energy consumption and sensor unrelia- bility. While GPS provides highly accurate location estimates, it has several limitations: some phones don't have GPS at all, the GPS sensor doesn't work in "urban canyons" (tall buildings and tunnels) or when the phone is inside a pocket, and the GPS on many phones is power-hungry and drains the battery quickly. In these cases, VTrack can use alter- native, less energy-hungry but noisier sensors like WiFi to estimate both a user's trajectory and travel time along the route. VTrack uses a hidden Markov model (HMM)-based map matching scheme and travel time estimation method that interpolates sparse data to identify the most probable road segments driven by the user and to attribute travel times to those segments. We present experimental results from real drive data and WiFi access point sightings gathered from a de- ployment on several cars. We show that VTrack can tolerate significant noise and outages in these location estimates, and still successfully identify delay-prone segments, and provide accurate enough delays for delay-aware routing algorithms. We also study the best sampling strategies for WiFi and GPS sensors for different energy cost regimes.
Conference Paper
Full-text available
We present AutoWitness, a system to deter, detect, and track personal property theft, improve historically dismal stolen property recovery rates, and disrupt stolen property distribution networks. A property owner embeds a small tag inside the asset to be protected, where the tag lies dormant until it detects vehicular movement. Once moved, the tag uses inertial sensor-based dead reckoning to estimate position changes, but to reduce integration errors, the relative position is reset whenever the sensors indicate the vehicle has stopped. The sequence of movements, stops, and turns are logged in compact form and eventually transferred to a server using a cellular modem after both sufficient time has passed (to avoid detection) and RF power is detectable (hinting cellular access may be available). Eventually, the trajectory data are sent to a server which attempts to match a path to the observations. The algorithm uses a Hidden Markov Model of city streets and Viterbi decoding to estimate the most likely path. The proposed design leverages low-power radios and inertial sensors, is immune to intransit cloaking, and supports post hoc path reconstruction. Our prototype demonstrates technical viability of the design; the volume market forces driving machine-to-machine communications will soon make the design economically viable.
Conference Paper
Full-text available
A growing number of mobile computing applications are cen- tered around the user's location. The notion of location is broad, ranging from physical coordinates (latitude/longitude) to logical labels (like Starbucks, McDonalds). While extensive research has been performed in physical localization, there have been few attempts in recognizing logical locations. This paper argues that the increasing number of sensors on mo- bile phones presents new opportunities for logical localiza- tion. We postulate that ambient sound, light, and color in a place convey a photo-acoustic signature that can be sensed by the phone's camera and microphone. In-built accelerometers in some phones may also be useful in inferring broad classes of user-motion, often dictated by the nature of the place. By combining these optical, acoustic, and motion attributes, it may be feasible to construct an identifiable fingerprint for logical localization. Hence, users in adjacent stores can be separated logically, even when their physical positions are ex- tremely close. We propose SurroundSense, a mobile phone based system that explores logical localization via ambience fingerprinting. Evaluation results from 51 different stores show that SurroundSense can achieve an average accuracy of 87% when all sensing modalities are employed. We be- lieve this is an encouraging result, opening new possibilities in indoor localization.
Conference Paper
Full-text available
Finding a person in a public place, such as in a library, confer- ence hotel, or shopping mall, can be difficult. The difficulty arises from not knowing where the person may be at that time; even if known, navigating through an unfamiliar place may be frustrating. Maps and floor plans help in some occa- sions, but such maps may not be always handy. In a small scale poll, 80% of users responded that the ideal solution would be "to have an escort walk me to the desired person". This paper identifies the possibility of using mobile phone sensors and opportunistic user-intersections to develop an elec- tronic escort service. By periodically learning the walking trails of different individuals, as well as how they encounter each other in space-time, a route can be computed between any pair of persons. The problem bears resemblance to rout- ing packets in delay tolerant networks, however, its applica- tion in the context of human localization raises distinct re- search challenges. We design and implement Escort, a system that guides a user to the vicinity of a desired person in a public place. We only use an audio beacon, randomly placed in the building, to enable a reference frame. We do not rely on GPS, WiFi, or war-driving to locate a person - the Escort user only needs to follow an arrow displayed on the phone. Evaluation results from experiments in parking lots and university build- ings show that, on average, the user is brought to within 8m of the destination. We believe this is an encouraging result, opening new possibilities in mobile, social localization.
Conference Paper
Full-text available
WiFi localization, the task of determining the phys- ical location of a mobile device from wireless sig- nal strengths, has been shown to be an accurate method of indoor and outdoor localization and a powerful building block for location-aware appli- cations. However, most localization techniques re- quire a training set of signal strength readings la- beled against a ground truth location map, which is prohibitive to collect and maintain as maps grow large. In this paper we propose a novel tech- nique for solving the WiFi SLAM problem using the Gaussian Process Latent Variable Model (GP- LVM) to determine the latent-space locations of un- labeled signal strength data. We show how GP- LVM, in combination with an appropriate motion dynamics model, can be used to reconstruct a topo- logical connectivity graph from a signal strength sequence which, in combination with the learned Gaussian Process signal strength model, can be used to perform efficient localization.
Conference Paper
Full-text available
For professional users such as firefighters and other first responders, GNSS positioning technology (GPS, assisted GPS) can satisfy outdoor positioning requirements in many instances. However, there is still a need for high-performance deep indoor positioning for use by these same professional users. This need has already been clearly expressed by various communities of end users in the context of WearIT@Work, an R&D project funded by the European Community's Sixth Framework Program. It is known that map matching can help for indoor pedestrian navigation. In most previous research, it was assumed that detailed building plans are available. However, in many emergency / rescue scenarios, only very limited building plan information may be at hand. For example a building outline might be obtained from aerial photographs or cataster databases. Alternatively, an escape plan posted at the entrances to many building would yield only approximate exit door and stairwell locations as well as hallway and room orientation. What is not known is how much map information is really required for a USAR mission and how much each level of map detail might help to improve positioning accuracy. Obviously, the geometry of the building and the course through will be factors consider. The purpose of this paper is to show how a previously published Backtracking Particle Filter (BPF) can be combined with different levels of building plan detail to improve PDR performance. A new in/out scenario that might be typical of a reconnaissance mission during a fire in a two-story office building was evaluated. Using only external wall information, the new scenario yields positioning performance (2.56 m mean 2D error) that is greatly superior to the PDR-only, no map base case (7.74 m mean 2D error). This result has a substantial practical significance since this level of building plan detail could be quickly and easily generated in many emergency instances. The technique could be used to mitigate- - heading errors that result from exposing the IMU to extreme operating conditions. It is hoped that this mitigating effect will also occur for more irregular paths and in larger traversed spaces such as parking garages and warehouses.
Conference Paper
We present Walkie-Markie - an indoor pathway mapping system that can automatically reconstruct internal pathway maps of buildings without any a-priori knowledge about the building, such as the floor plan or access point locations. Central to Walkie-Markie is a novel exploitation of the WiFi infrastructure to define landmarks (WiFi-Marks) to fuse crowdsourced user trajectories obtained from inertial sensors on users' mobile phones. WiFi-Marks are special pathway locations at which the trend of the received WiFi signal strength changes from increasing to decreasing when moving along the pathway. By embedding these WiFi-Marks in a 2D plane using a newly devised algorithm and connecting them with calibrated user trajectories, Walkie-Markie is able to infer pathway maps with high accuracy. Our experiments demonstrate that Walkie-Markie is able to reconstruct a high-quality pathway map for a real office-building floor after only 5-6 rounds of walks, with accuracy gradually improving as more user data becomes available. The maximum discrepancy between the inferred pathway map and the real one is within 3m and 2.8m for the anchor nodes and path segments, respectively.
Conference Paper
As industrial and academic communities become increasingly interested in Indoor Positioning Systems (IPSs), a plethora of technologies are gaining maturity and competing for adoption in the global smartphone market. In the near future, we expect busy places, such as schools, airports, hospitals and large businesses, to be outfitted with multiple IPS infrastructures, which need to coexist, collaborate and / or compete for users. In this paper, we examine the novel problem of estimating the accuracy of co-located positioning systems, and selecting which one to use where. This is challenging because 1) we do not possess knowledge of the ground truth, which makes it difficult to empirically estimate the accuracy of an indoor positioning system; and 2) the accuracy reported by a positioning system is not always a faithful representation of the real accuracy. In order to address these challenges, we model the process of a user moving in an indoor environment as a Hidden Markov Model (HMM), and augment the model to take into account vector (instead of scalar) observations, and prior knowledge about user mobility drawn from personal electronic calendars. We then propose an extension of the Baum-Welch algorithm to learn the parameters of the augmented HMM. The proposed HMM-based approach to learning the accuracy of indoor positioning systems is validated and tested against competing approaches in several real-world indoor settings.
Conference Paper
Due to the complementary nature of inertial navigation system (INS) and Wi-Fi positioning principles, an INS/Wi-Fi integrated system is expected to form a low-cost and continuous indoor navigation solution with better performance than using the standalone systems. In this paper, we explore the integration of Wi-Fi measurements with data from microelectromechanical systems (MEMS) based inertial measurement unit (IMU) for indoor vehicle navigation. Two enhancements, which employ adaptive Kalman filtering (AKF) and vehicle constraints, for supporting the integrated system are presented. One field experiment has been conducted for estimating the trajectory of a mobile robot vehicle. The numerical results show that the enhanced integrated system provides higher navigation accuracy, compared to using standalone Wi-Fi positioning and conventional INS/Wi-Fi integration.
Article
Radio Frequency (RF) fingerprinting, based onWiFi or cellular signals, has been a popular approach to indoor localization. However, its adoption in the real world has been stymied by the need for sitespecific calibration, i.e., the creation of a training data set comprising WiFi measurements at known locations in the space of interest. While efforts have been made to reduce this calibration effort using modeling, the need for measurements from known locations still remains a bottleneck. In this paper, we present Zee -- a system that makes the calibration zero-effort, by enabling training data to be crowdsourced without any explicit effort on the part of users. Zee leverages the inertial sensors (e.g., accelerometer, compass, gyroscope) present in the mobile devices such as smartphones carried by users, to track them as they traverse an indoor environment, while simultaneously performing WiFi scans. Zee is designed to run in the background on a device without requiring any explicit user participation. The only site-specific input that Zee depends on is a map showing the pathways (e.g., hallways) and barriers (e.g., walls). A significant challenge that Zee surmounts is to track users without any a priori, user-specific knowledge such as the user's initial location, stride-length, or phone placement. Zee employs a suite of novel techniques to infer location over time: (a) placement-independent step counting and orientation estimation, (b) augmented particle filtering to simultaneously estimate location and user-specific walk characteristics such as the stride length,(c) back propagation to go back and improve the accuracy of ocalization in the past, and (d) WiFi-based particle initialization to enable faster convergence. We present an evaluation of Zee in a large office building.
Article
Indoor localization is of great importance for a range of pervasive applications, attracting many research efforts in the past decades. Most radio-based solutions require a process of site survey, in which radio signatures of an interested area are annotated with their real recorded locations. Site survey involves intensive costs on manpower and time, limiting the applicable buildings of wireless localization worldwide. In this study, we investigate novel sensors integrated in modern mobile phones and leverage user motions to construct the radio map of a floor plan, which is previously obtained only by site survey. On this basis, we design LiFS, an indoor localization system based on off-the-shelf WiFi infrastructure and mobile phones. LiFS is deployed in an office building covering over 1600m2, and its deployment is easy and rapid since little human intervention is needed. In LiFS, the calibration of fingerprints is crowdsourced and automatic. Experiment results show that LiFS achieves comparable location accuracy to previous approaches even without site survey.
Article
This paper addresses the problem of tracking a group of mobile sensors in an environment where there is intermittent or no access to a localization service, such as the Global Positioning System. Example applications include tracking personnel underground or animals under dense tree canopies. We assume that each sensor uses inertial, visual or mechanical odometry to measure its relative movement as a series of displacement vectors. Each displacement vector suffers a small quantity of error which compounds, causing the overall accuracy of the positional estimate to decrease with time. The primary contribution of this paper is a novel offline method of counteracting this error by exploiting opportunistic radio encounters between sensors. We fuse encounter information with the displacement vectors to build a graph that models sensor mobility. We show that two dimensional sensor tracking is equivalent to finding an embedding of this graph in the plane. Finally, using radio, inertial and ground truth trace data, we conduct simulations to observe how the number of anchors, transmission range and radio noise affect the performance of the proposed model. We compare these results to those from a competing model in the literature.
Article
We present AutoWitness, a system to deter, detect, and track personal property theft, improve historically dismal stolen property recovery rates, and disrupt stolen property distribution networks. A property owner embeds a small tag inside the asset to be protected, where the tag lies dormant until it detects vehicular movement. Once moved, the tag uses inertial sensor-based dead reckoning to estimate position changes, but to reduce integration errors, the relative position is reset whenever the sensors indicate the vehicle has stopped. The sequence of movements, stops, and turns are logged in compact form and eventually transferred to a server using a cellular modem after both sufficient time has passed (to avoid detection) and RF power is detectable (hinting cellular access may be available). Eventually, the trajectory data are sent to a server which attempts to match a path to the observations. The algorithm uses a Hidden Markov Model of city streets and Viterbi decoding to estimate the most likely path. The proposed design leverages low-power radios and inertial sensors, is immune to intransit cloaking, and supports post hoc path reconstruction. Our prototype demonstrates technical viability of the design; the volume market forces driving machine-to-machine communications will soon make the design economically viable.
Conference Paper
Orientation estimation can be executed by comparing the output of a 3D accelerometer and a 3D magnetometer with respectively gravity and local magnetic field vectors. For this purpose, an unscented Kalman filter was designed and tested. However, accelerometers also measure motion other than gravity, resulting in an error when estimating orientation directly from their output signals. Therefore, extra filters are added and the input parameters of the Kalman filter are dynamically varied, in order to reduce the effect of motion. Simulations are performed to tune the filter parameters for minimal motion influence without hampering actual orientation tracking. Satisfactory orientation tracking is performed with the filter using actual sensor nodes.
Conference Paper
In this paper we present an extension to odometry based SLAM for pedestrians that incorporates human-reported measurements of recognizable features, or “places” in an environment. The method which we have called “PlaceSLAM” builds on the Simultaneous Localization and Mapping (SLAM) principle in that a spatial representation of such places can be built up during the localization process. We see an important application to be in mapping of new areas by volunteering pedestrians themselves, in particular to improve the accuracy of “FootSLAM” which is based on human step estimation (odometry). We present a description of various flavors of PlaceSLAM and derive a Bayesian formulation and particle filtering implementation for the most general variant. In particular we distinguish between two important cases which depend on whether the pedestrian is required to report a place's identifier or not. Our results based on experimental data show that our approach can significantly improve the accuracy and stability of FootSLAM and this with very little additional complexity. After mapping has been performed, users of such improved FootSLAM maps need not report places themselves.
Conference Paper
The problem of matching measured latitude/longitude points to roads is becoming increasingly important. This paper describes a novel, principled map matching algorithm that uses a Hidden Markov Model (HMM) to find the most likely road route represented by a time-stamped sequence of latitude/longitude pairs. The HMM elegantly accounts for measurement noise and the layout of the road network. We test our algorithm on ground truth data collected from a GPS receiver in a vehicle. Our test shows how the algorithm breaks down as the sampling rate of the GPS is reduced. We also test the effect of increasing amounts of additional measurement noise in order to assess how well our algorithm could deal with the inaccuracies of other location measurement systems, such as those based on WiFi and cell tower multilateration. We provide our GPS data and road network representation as a standard test set for other researchers to use in their map matching work.
Conference Paper
Location information is an important source of context for ubiquitous computing systems. This paper looks at how a foot-mounted inertial unit, a detailed building model, and a particle filter can be combined to provide absolute posi- tioning, despite the presence of drift in the inertial unit and without knowledge of the user's initial location. We show how to handle multiple floors and stairways, how to handle symmetry in the environment,and how to initialise the local- isation algorithm using WiFi signal strength to reduce initial complexity. We evaluate the entire system experimentally,using an inde- pendent tracking system for ground truth. Our results show that we can track a user throughouta 8725 m2 buildingspan- ning three floors to within 0.5 m 75% of the time, and to within 0.73 m 95% of the time.
Conference Paper
Location awareness is an important capability for mobile computing. Yet inexpensive, pervasive positioning—a requirement for wide-scale adoption of location-aware computing—has been elusive. We demonstrate a radio beacon-based approach to location, called Place Lab, that can overcome the lack of ubiquity and high-cost found in existing location sensing approaches. Using Place Lab, commodity laptops, PDAs and cell phones estimate their position by listening for the cell IDs of fixed radio beacons, such as wireless access points, and referencing the beacons' positions in a cached database. We present experimental results showing that 802.11 and GSM beacons are sufficiently pervasive in the greater Seattle area to achieve 20-30 meter median accuracy with nearly 100% coverage measured by availability in people's daily lives.
Article
This paper is concerned with orientation estimation using inertial and magnetic sensors. A quaternion-based indirect Kalman filter structure is used. The magnetic sensor output is only used for yaw angle estimation using two-step measurement updates. External acceleration is estimated from the residual of the filter and compensated by increasing the measurement noise covariance. Using the direction information of external information, the proposed method prevents unnecessarily increasing the measurement noise covariance corresponding to the accelerometer output, which is not affected by external acceleration. Through numerical examples, the proposed method is verified.
Article
Heterogeneous wireless clients measure signal strength differently. This is a fundamental problem for indoor location fingerprinting, and it has a high impact on the positioning accuracy. Mapping-based solutions have been presented that require manual and error-prone calibration for each new client. This article presents hyperbolic location fingerprinting, which records fingerprints as signal strength ratios between pairs of base stations instead of absolute signal strength values. This article also presents an automatic mapping-based method that avoids calibration by learning from online measurements. The evaluation shows that the solutions can address the signal strength heterogeneity problem without requiring extra manual calibration.
Article
A navigation system that tracks the location of a person on foot is useful for finding and rescuing firefighters or other emergency first responders, or for location-aware computing, personal navigation assistance, mobile 3D audio, and mixed or augmented reality applications. One of the main obstacles to the real-world deployment of location-sensitive wearable computing, including mixed reality (MR), is that current position-tracking technologies require an instrumented, marked, or premapped environment. At InterSense, we've developed a system called NavShoe, which uses a new approach to position tracking based on inertial sensing. Our wireless inertial sensor is small enough to easily tuck into the shoelaces, and sufficiently low power to run all day on a small battery. Although it can't be used alone for precise registration of close-range objects, in outdoor applications augmenting distant objects, a user would barely notice the NavShoe's meter-level error combined with any error in the head's assumed location relative to the foot. NavShoe can greatly reduce the database search space for computer vision, making it much simpler and more robust. The NavShoe device provides not only robust approximate position, but also an extremely accurate orientation tracker on the foot.
Conference Paper
The proliferation of mobile computing devices and local-area wireless networks has fostered a growing interest in location-aware systems and services. In this paper we present RADAR, a radio-frequency (RF)-based system for locating and tracking users inside buildings. RADAR operates by recording and processing signal strength information at multiple base stations positioned to provide overlapping coverage in the area of interest. It combines empirical measurements with signal propagation modeling to determine user location and thereby enable location-aware services and applications. We present experimental results that demonstrate the ability of RADAR to estimate user location with a high degree of accuracy
Article
Increasingly, for many application areas, it is becoming important to include elements of nonlinearity and non-Gaussianity in order to model accurately the underlying dynamics of a physical system. Moreover, it is typically crucial to process data on-line as it arrives, both from the point of view of storage costs as well as for rapid adaptation to changing signal characteristics. In this paper, we review both optimal and suboptimal Bayesian algorithms for nonlinear/non-Gaussian tracking problems, with a focus on particle filters. Particle filters are sequential Monte Carlo methods based on point mass (or "particle") representations of probability densities, which can be applied to any state-space model and which generalize the traditional Kalman filtering methods. Several variants of the particle filter such as SIR, ASIR, and RPF are introduced within a generic framework of the sequential importance sampling (SIS) algorithm. These are discussed and compared with the standard EKF through an illustrative example
Article
We present conditional random fields, a framework for building probabilistic models to segment and label sequence data. Conditional random fields offer several advantages over hidden Markov models and stochastic grammars for such tasks, including the ability to relax strong independence assumptions made in those models. Conditional random fields also avoid a fundamental limitation of maximum entropy Markov models (MEMMs) and other discriminative Markov models based on directed graphical models, which can be biased towards states with few successor states. We present iterative parameter estimation algorithms for conditional random fields and compare the performance of the resulting models to HMMs and MEMMs on synthetic and natural-language data.
3d orientation tracking based on unscented kalman filtering of accelerometer and magnetometer data
  • B Huyghe
  • J Doutreloigne
  • J Vanfieteren