An Effective GNSS/PDR Fusion Positioning Algorithm on Smartphones for Challenging Scenarios

Jingkui Zhang,Baoguo Yu,Yuxiang Ge,Jingxiang Gao,Chuanzhen Sheng
DOI: https://doi.org/10.3390/s24051452
IF: 3.9
2024-02-24
Sensors
Abstract:The location-based smartphone service brings new development opportunities for seamless indoor/outdoor positioning. However, in complex scenarios such as cities, tunnels, overpasses, forests, etc., using only GNSS on smartphones cannot provide stable and reliable positioning results. Usually, additional sensors are needed to assist GNSS. This paper investigates the GNSS positioning algorithm assisted by pedestrian dead reckoning (PDR) in complex scenarios. First, we introduce a step detection algorithm based on the peak–valley of acceleration modulus, and the Weinberg model and the Mahony algorithm in PDR are used to estimate step length and heading. On this basis, we evaluated the performance of GNSS/PDR fusion positioning in an open scenario, a semiopen scenario, and a blocked scenario, respectively. Finally, we develop a GNSS/PDR real-time positioning software, called China University of Mining and Technology-POSitioning (CUMT-POS) version 1.0, on the Android 10 platform. By comparing GNSS solutions, PDR solutions, GNSS/PDR solutions, and real-time kinematic (RTK) solutions, we verify the potential auxiliary ability of PDR for GNSS positioning in complex environments, proving that multisource sensor fusion positioning significantly improves reliability and stability. Our research can help the realization of urban informatization and smart cities.
engineering, electrical & electronic,chemistry, analytical,instruments & instrumentation
What problem does this paper attempt to address?
The problem that this paper attempts to solve is that in complex scenarios, using only GNSS on smart phones cannot provide stable and reliable positioning results. Specifically, the paper focuses on situations where GNSS signals are unstable or unavailable due to signal blockage or multipath effects in complex environments such as cities, tunnels, overpasses, and forests. In order to improve the positioning accuracy and stability in these environments, the paper studies the GNSS/PDR fusion positioning algorithm combined with Pedestrian Dead - Reckoning (PDR). ### Background and Problem Description of the Paper 1. **Limitations of GNSS**: - GNSS can provide high - precision absolute position information, but its performance is limited by whether the signal is unobstructed. In scenarios where the signal is blocked or there are severe multipath effects, the reduction in the number of received satellites and the deterioration of the observation quality will lead to a significant decline in positioning accuracy. 2. **Advantages and Challenges of PDR**: - PDR provides relative position information through inertial sensors (such as accelerometers and gyroscopes). Its positioning accuracy is not affected by external factors, but errors will accumulate over a long period. 3. **Requirement for Fusion Positioning**: - To overcome the limitations of a single technology, the paper proposes to combine GNSS and PDR, using the absolute position information of GNSS and the relative position information of PDR to improve the positioning accuracy and robustness in complex environments. ### Main Contributions of the Paper 1. **Gait Detection Algorithm**: - Based on the peak - valley characteristics of the acceleration magnitude, a gait detection algorithm is proposed, and the Butterworth filter is used to denoise the acceleration data, which improves the accuracy of gait detection. 2. **Stride Length Estimation Model**: - The Weinberg model is adopted to estimate the stride length. This model is based on acceleration parameters and can relatively accurately estimate the stride lengths of different pedestrians and in different motion states. 3. **Heading Estimation Method**: - The Mahony algorithm combined with the complementary filter is used to accurately estimate the heading of pedestrians through the data of accelerometers, magnetometers and gyroscopes. 4. **GNSS/PDR Fusion Positioning Algorithm**: - A GNSS/PDR fusion positioning algorithm based on Kalman filtering is designed, which combines the absolute position information provided by GNSS and the relative position information provided by PDR, improving the positioning accuracy and stability. ### Experimental Verification The paper conducted experiments in open, semi - open and occluded scenarios respectively to verify the effectiveness of the GNSS/PDR fusion positioning algorithm. The experimental results show that the fusion positioning algorithm is significantly superior to using GNSS or PDR alone in terms of positioning accuracy and stability in complex environments. ### Conclusion By combining GNSS and PDR, this paper proposes an effective fusion positioning algorithm, which solves the problem of unstable smartphone positioning in complex environments and provides technical support for the realization of urban informatization and smart cities.