Accurate positioning is critical to Intelligent Transportation Systems (ITSs). Current research primarily focuses on improving Global Navigation Satellite System (GNSS) positioning accuracy and continuity through multi-sensor integration. With the emergence of new industries such as assisted driving, the credibility of positioning results has gradually attracted attention. To explore the feasibility of achieving credible positioning, this paper presents an optimized Inertial Measurement Unit (IMU) and camera tightly augmented GNSS Real Time Kinematic (RTK) model, along with the positioning credibility assessment. In this model, multi-factors that affect the positioning errors are considered as the feature inputs of the Convolutional Neural Network-Long Short-Term Memory (CNN-LSTM) network, then a credible factor and its uncertainty are generated. Moreover, a positioning optimization algorithm is presented based on the credible factor. To evaluate the effectiveness of the presented model, several sets of vehicle-borne data in urban environments are processed and analyzed. Results illustrate that (1) the presented positioning model achieves comparable positioning and superior orientation determination accuracy compared to existing state-of-the-art methods; (2) the generated credible factor can envelop 94% horizontal positioning errors and 84% vertical positioning errors with envelope levels of 5 cm and 7cm; (3) the error coverage rate of confidence interval generated by the uncertainty of credible factor in horizontal and vertical directions can reach 94% and 87.57%, which is close to the theoretically set 95% confidence interval for horizontal direction; (4) the positioning results can be optimized while applying the position optimization algorithm based on credible factor.

Smartphone-based pedestrian dead reckoning (PDR) is widely used in Internet of Things (IoT) applications. However, the accuracy and reliability of PDR could be affected by the users’ environments significantly. To upgrade PDR’s performance, we present an enhanced PDR algorithm, in which pedestrian motion constraints, smartphone sensors, and a combined step detection method are integrated with PDR to provide continuous, accurate, and reliable position solutions. In such a method, measurements of triaxis accelerometers, triaxis gyroscopes, triaxis magnetometers, a barometer, and a global navigation satellite system (GNSS) chip from Huawei Mate30Pro are integrated by an extended Kalman filter (EKF). Pedestrian motions like motionless and linear motion form the constraints to upgrade the performance of the multisensor enhanced PDR. Results based on a set of experimental data demonstrated that the proposed PDR could provide positioning accuracy in terms of the root mean-square error (RMSE) within 1.5 m. Compared to the PDR-only, the position enhancements on vertical and horizontal from a barometer, triaxis accelerometers, triaxis magnetometers, GNSS, and motion constraints are visible with improvement percentages of more than 98.9%, which makes the position solutions of the presented PDR much more reliable than the conventional PDR.

The requirement for credible and reliable positioning of a multi-sensor integration system is an essential foundation for comprehensive PNT (Positioning, Navigation, and Timing) services. However, the credibility of positioning results based on multi-sensor integration is difficult to evaluate and even there is no effective mode at present. To try to solve such a problem, a credible positioning model based on the tight integration of BDS Real Time Kinematic (RTK) and Inertial Navigation System (INS) is presented in this paper. In such a model, a multi-information cross-validation algorithm is introduced to ensure the credibility of RTK/INS tight integration. Meanwhile, a backward quality checking model is generated based on the result of the calculated credible positioning error level, which can optimize the positioning results of RTK/INS integration furtherly. After the mathematical model descriptions, a simulated test and a real experiment are adopted to evaluate this presented method. Results illustrated that: 1) envelope level of the credible factor for the real error can reach the centimeter level and the average probability of unenveloped error is within 2%; 2) after applying the quality control scheme based on credible positioning feedback information, the positioning results of RTK/INS integration are improved by 67.6%, 78.9%, and 77.3% on average.

Real-time kinematic (RTK) based on single-frequency observation is still widely used in many fields due to the characteristics of low-cost and low-power consumption. However, the positioning performance of single-frequency RTK in terms of accuracy, stability, and continuity would be significantly degraded during the harsh satellite environments. To improve the performance, this contribution presented a model of multisensor and analytical observations augmented the single-frequency RTK tightly based on a modified Psi-angle state model. In such a model, the single frequency observations of the new signal of third generation BDS (BDS-3) B2b are tightly integrated with inertial measurements, odometer data, dual-antenna attitude, and nonholonomic constraint (NHC). To evaluate the presented model, the typical navigation performance and the ambiguity resolution (AR) performance are analyzed based on a set of vehicle-borne data. Results illustrated that the inertial navigation system (INS) would bring about 13.5%, 16.2%, and 12.3% position enhancements to the BDS-3 B2b RTK mode. Such improvements could be up to 15.9%, 16.2%, and 25.2% while adding the NHC and odometer data. Besides, NHC and odometer also upgrade the attitude accuracy visibly in pitch and heading directions, with enhancements of about 16.9% and 62.9%. In contrast, augmentations from the dual-antenna attitude are mainly presented in terms of heading angle with about 29.5% compared to the RTK/INS/Odometer/NHC tight integration mode. Besides, the convergence time of yaw angle is visibly enhanced while using the odometer/NHC, the dual-antenna heading, or the two together. Moreover, the AR performance could also be improved while using the presented model. Due to the enhancements in position and attitude brought by different sensors, the ADOP performance would be improved to varying degrees. Besides, the fixed rate and reliability of AR could also be enhanced.

The third generation BeiDou Navigation Satellite System (BDS-3) can provide high-accuracy service for vehicle-borne positioning and navigation. The performance of BDS-3 in terms of accuracy, continuity, and reliability, however, are significantly degraded in environments with weak satellite observability. To improve the positioning performance of BDS-3 around the partial and complete signal-blocked areas, this paper presents a multi-sensor and analytical constraints tightly augmented BDS-3 Real-time Kinematic (RTK). The Non-holonomic Constraint (NHC), vehicle-odometer, and dual-antenna-based attitude constraints applied to restrain the drift of the position estimate from Inertial Measurement Units (IMU). Compared to previous work, it’s the first time to reveal that the Ambiguity Resolution (AR) performance of RTK can be augmented by integrating these measurements. To validate the performance of this proposed method, the drifts of position and attitude during the poor satellite observability and the impacts of those sensors and constraints on ambiguity fixing are analyzed based on an urban vehicle-borne test. The test results illustrate that the presented model brought 52.73%, 55.56%, and 49.44% positioning accuracy improvements in the north, east, and vertical components compared to the RTK mode. Meanwhile, 10.87% and 57.20% attitude accuracy enhancements in roll and heading were obtained compared to the traditional RTK/IMU tight integration mode. In addition, this presented model can retain the accuracy of position and attitude during the partial and complete satellite signal outage periods, and can also improve the ambiguity resolution performance significantly.

Conventional indoor positioning techniques (i.e., Wi-Fi, Bluetooth, and ZigBee) have been well researched and widely used for the indoor localization service. However, these methodologies can hardly provide users high-accuracy positioning solution. Along with the rapid developments of the internet of things (IoT), the conventional indoor positioning methods cannot match the high-accuracy requirements of IoT. In this paper, we present a practical high-accuracy indoor positioning algorithm, in which the Two-way Time-of-arrival (TWTOA) ranging-based ultra-wide-band (UWB) is integrated tightly with the inertial navigation system (INS). In such a method, the INS is utilized to overcome the impacts of the non-line-of-sight (NLOS) delays on UWB high-accuracy positioning. According to the experiment results, UWB only can provide decimeter-level positioning accuracy. However, UWB positioning accuracy would degrade significantly while suffering NLOS delays. After integrating these NLOS-contaminated UWB ranges with INS tightly, the influences of NLOS could be inhibited, which results in higher accuracy positioning results. Meanwhile, the field test also demonstrated that the number of available UWB base stations could also present visible impacts on the positioning accuracy of the UWB/INS tightly coupled integration.

Currently, the Global Navigation Satellite System (GNSS) measurements-based Real-Time Kinematic (RTK) and Precise Point Positioning (PPP) technologies are the two effective methods to provide users with centimeter-level positioning solutions in strong satellite observability environments. However, the performance (precision, continuity, and reliability) would seriously degrade while suffering challenging environments. To improve GNSS performance around the signals blocked areas, the Inertial Navigation System (INS) and vision camera sensors are integrated with GNSS in this contribution. Firstly, we design a GNSS position/INS/Vision integration based on Multi-State Constraint Kalman Filter (MSCKF). Then, a set of vehicle-borne data collected under GNSS-denied environments are processed and analyzed to assess the performance of such presented algorithm. Results illustrate that (1) In GNSS-denied environments, solutions calculated by both PPP and RTK are un-continuous; (2) however, its performance can be improved while using multi-GNSS observations; (3) with the aids from INS and vision, GNSS performance in terms of accuracy, continuity, and availability can be upgraded significantly even suffering satellite signal outages.

The inertial navigation system (INS) has been widely used to provide self-contained and continuous motion estimation in intelligent transportation systems. Recently, the emergence of chip-level inertial sensors has expanded the relevant applications from positioning, navigation, and mobile mapping to location-based services, unmanned systems, and transportation big data. Meanwhile, benefit from the emergence of big data and the improvement of algorithms and computing power, machine learning (ML) has become a consensus tool that has been successfully applied in various fields. This article reviews the research on using ML technology to enhance inertial sensing from various aspects, including sensor design and selection, calibration and error modeling, navigation and motion-sensing algorithms, multi-sensor information fusion, system evaluation, and practical application. It summarizes the state of the art, advantages, and challenges on each aspect, and points out future research directions.

Localization techniques are becoming key to add location context to the Internet-of-Things (IoT) data without human perception and intervention. Meanwhile, the newly emerged low-power wide-area network (LPWAN) and 5G technologies have become strong candidates for mass-market localization applications. However, various error sources have limited localization performance by using such IoT signals. This article reviews the IoT localization system through the following sequence: IoT localization system review, localization data sources, localization algorithms, localization error sources and mitigation, and localization performance evaluation. Compared to the related surveys, this article has a more comprehensive and state-of-the-art review on IoT localization methods, an original review on IoT localization error sources and mitigation, an original review on IoT localization performance evaluation, and a more comprehensive review of IoT localization applications, opportunities, and challenges. Thus, this survey provides comprehensive guidance for peers who are interested in enabling localization ability in the existing IoT systems, using IoT systems for localization, or integrating IoT signals with the existing localization sensors.

Location is key to spatialize Internet of Things (IoT) data. However, it is challenging to use low-cost IoT devices for robust unsupervised localization (i.e., localization without training data that have known location labels). Thus, this article proposes a deep-reinforcement-learning (DRL)-based unsupervised wireless-localization method. The main contributions are as follows: 1) this article proposes an approach to model a continuous wireless-localization process as a Markov decision process and process it within a DRL framework; 2) to alleviate the challenge of obtaining rewards when using unlabeled data (e.g., daily life crowdsourced data), this article presents a reward-setting mechanism, which extracts robust landmark data from unlabeled wireless received signal strengths (RSS); and 3) to ease requirements for model retraining when using DRL for localization, this article uses RSS measurements together with agent location to construct DRL inputs. The proposed method is tested by using field testing data from multiple Bluetooth 5 smart ear tags in a pasture. Meanwhile, the experimental verification process reflects the advantages and challenges for using DRL in wireless localization.

The next-generation Internet of Things (IoT) systems have an increasingly demand on intelligent localization which can scale with big data without human perception. Thus, traditional localization solutions without accuracy metric will greatly limit vast applications. Crowd sourcing-based localization has been proven to be effective for mass-market location-based IoT applications. This paper proposes an enhanced crowd sourcing-based localization method by integrating inertial, wireless, and magnetic sensors. Both wireless and magnetic fingerprinting accuracy are predicted in real time through the introduction of fingerprinting accuracy indicators (FAIs) from three levels (i.e., signal, geometry, and database). The advantages and limitations of these FAI factors and their performances on predicting location errors and outliers are investigated. Furthermore, the FAI-enhanced extended Kalman filter (EKF) is proposed, which improved the dead-reckoning (DR)/WiFi, DR/Magnetic, and DR/WiFi/Magnetic integrated localization accuracy by 30.2%, 19.4%, and 29.0%, and reduced the maximum location errors by 41.2%, 28.4%, and 44.2%, respectively. These outcomes confirm the effectiveness of the FAI-enhanced EKF on improving both accuracy and reliability of multisensor integrated localization using crowd sourced data.