With the rise of location services applications, the need for navigation and positioning extends from isolated environments to a seamless, full domain. Aiming at the problem of low positioning accuracy in the transition area between indoor and outdoor, and it is difficult to achieve smooth switching, this article proposes an indoor and outdoor seamless positioning system for smartphones based on environment perception. Firstly, based on the hidden Markov model (HMM), we use smartphone observation data to select optimal features to construct an environmental perception model. Secondly, by increasing the magnetic field strength as an observation feature, the time delay for switching between different environments is shortened. Finally, based on the environmental perception results, a seamless indoor and outdoor positioning system is constructed. In the transition area between indoor and outdoor, to ensure the reliability of the positioning system, a Global Navigation Satellite System/ Inertial Navigation System/ Bluetooth Low Energy (GNSS/INS/BLE) integrated positioning model is proposed. In the outdoor and outdoor transition areas, we refine the stochastic model, which significantly improves the accuracy. To comprehensively evaluate the system performance, experiments are carried out in a complex environment. The experimental results show that the proposed perception model can accurately identify the changes of the environment, and the accuracy of judgment reaches more than 95%. It can support the seamless positioning system to select appropriate positioning models in different environments to achieve accurate and continuous positioning. The maximum positioning error does not exceed 0.5m.

Smartphone positioning methods relying solely on global navigation satellite system (GNSS) are susceptible to significant errors in urban environments due to challenging observation conditions. To enhance position accuracy, this article introduces a novel fifth-generation (5G) observation model based on unit vectors and evaluates its performance using several user locations and precision observation data. The results demonstrate that the unit vector method, compared to the traditional angle of departure (AOD) approach, effectively reduces the nonlinearity of the observation model, leading to a substantial improvement in the 5G system’s positioning accuracy. Furthermore, we conduct experiments to investigate the impact of the geometric distribution of 5G base station (BS) height on positioning accuracy. The results reveal a significant influence, highlighting the importance of considering BS height in the positioning process. In the urban environment, introducing 5G BS angle observations as an aid to GNSS significantly enhances the reliability of smartphone positioning, achieving positioning errors of less than one meter. Notably, this approach outperforms the introduction of 5G distance observations, providing valuable insights for the development of seamless indoor and outdoor positioning systems in the future.

For long baseline in a network, the traditional combined ionosphere-free (IF) + wide-lane (WL) strategy is commonly used, but the residual tropospheric delays and larger noise hamper the basic ambiguity resolution (AR). With the completion of the BeiDou global navigation satellite system (BDS-3) and the quad-frequency signals provided by BDS-3 satellites, we can construct more combinations that are conducive to ambiguity resolution. Compared with ionosphere-free linear combinations, we estimated ionospheric delay using three independent WL observations, and formed an ionosphere-weighted model using uncombined code and phase observations, which proved to be quite effective. Based on the real quad-frequency BDS-3 observations of two CORS (Continuously Operating Reference Stations) and two user stations, we processed eight days of data to study the formal and empirical ambiguity success rates and user positioning errors. The rounding success rate of WL ambiguity was significantly improved with ionospheric correction. The success rate of the basic ambiguity increased from 94.4 and 96.1% to 98.0% using the quad-frequency ionosphere-weighted (QFIW) model compared with the double-frequency ionosphere-free (DFIF) model and the triple-frequency geometry-based (TFGB) model. Furthermore, the user E/N/U positioning accuracy improved by 20.6/31.5/13.1% and 6.3/22.9/5.8%, respectively.

The need for high-precision and dependable navigation and positioning services is rising in the mass market as a result of the BDS-3’s completion and the performance development of smartphone chips. However, compared with the geodesic receiver, the positioning accuracy of smartphones is severely degraded due to the low antenna gain and poor anti-interference ability. In March 2022, Redmi released K50 smartphones that support BDS tri-frequency, providing conditions for further improvement of smartphone positioning. First, we collected the multi-constellation and multi-frequency data of the K50 to evaluate the availability of tri-frequency observations. Then, we compared the observations between different frequencies and found that the pseudorange data quality of the L5/E5a is better than that of L1/E1. For the BDS, the pseudorange data quality of the B2a is obviously better than that of the other two frequencies. For carrier observations, expect for the initial phase biases of B1C, there is no significant difference in data quality between different frequencies. It means that the traditional random model cannot accurately reflect the pseudorange quality of smartphone observation. We used the zero-baseline double difference residual method to calculate the smartphone observation accuracy and refined the random model. Finally, we verified the performance of the refined model through location tests. The results show that the model refinement can significantly improve positioning accuracy and reliability in the single point positioning test and shorten the convergence time in relative positioning. In kinematic positioning, the improved model can obtain the positioning accuracy of a single epoch centimeter-level even when the signal is slightly blocked, which is of great significance for high-precision navigation of smartphones in urban environments.

Positioning methods that rely solely on global satellite navigation system (GNSS) measurements have significantly decreased reliability in urban environments and cannot meet the needs of industries, such as autonomous driving. This article introduces a novel fifth-generation (5G) joint time of arrival (TOA) and unit vector observation model that takes rotation into account, and evaluates its performance through multiple sets of simulated data. The results show that compared with the traditional joint TOA and angle of arrival (AOA) observation model, the proposed model can notably improve the positioning accuracy, and the effect is better after iteration. Furthermore, we conduct a new stochastic model for 5G measurements and investigate the impact of model coefficients on performance. The results reveal that inappropriate coefficients will lead to loss of observational information, highlighting the importance of stochastic models. In urban environments, the proposed 5G joint observation model can effectively assist GNSS and provide more accurate navigation services and rotation estimation, providing valuable insights for the future development of smart transportation.