Currently, indoor locations based on the received signal strength (RSS) of Wi-Fi are attracting more and more attention thanks to the technology’s low cost, low power consumption and wide availability in mobile devices. However, the accuracy of Wi-Fi positioning is limited, due to the signal fluctuation and indoor multipath interference. In order to overcome this problem, this paper proposes a robust and accurate Wi-Fi fingerprint location recognition method based on a deep neural network (DNN). A stacked denoising auto-encoder (SDAE) is used to extract robust features from noisy RSS to construct a feature-weighted fingerprint database offline. We use the combination of the weights of posteriori probability and geometric relationship of fingerprint points to calculate the coordinates of unknown points online. In addition, we use constrained Kalman filtering and hidden Markov models (HMM) to smooth and optimize positioning results and overcome the influence of gross error on positioning results, combined with characteristics of user movement in buildings, both dynamic and static. The experiment shows that the DNN is feasible for position recognition, and the method proposed in this paper is more accurate and stable than the commonly used Wi-Fi positioning methods in different scenes.

In the meteorology of Global Navigation Satellite System, the weighted mean temperature (Tm) is a key parameter in the process of converting the zenith wetness delay into precipitable water vapor, and it plays an important role in water vapor monitoring. In this research, two deep learning algorithms, namely, recurrent neural network (RNN) and long short-term memory neural network (LSTM), were used to build a high-precision weighted mean temperature model for China using their excellent time series memory capability. The model needs site location information and measured surface temperature to predict the weighted mean temperature. We used data from 118 stations in and around China provided by the Integrated Global Radiosonde Archive from 2010 to 2015 to train the model and data from 2016 for model testing. The root mean square error (RMSE) of the RNN_Tm and LSTM_Tm models were 3.01 K and 2.89 K, respectively. Compared with the values calculated by the empirical GPT3 model, the accuracy was improved by 31.1% (RNN_Tm) and 33.9% (LSTM_Tm). In addition, we selected another 10 evenly distributed stations in China and used the constructed model to test the prediction capability of the weighted mean temperature from 2010 to 2016. The RMSE values were 2.95 K and 2.86 K, which proved that the model also exhibits high generalization in non-modeling sites in China. In general, the RNN_Tm and LSTM_Tm models have a good performance in weighted mean temperature prediction.

Multi-constellation global navigation satellite systems (GNSSs) are expected to enhance the capability of precise point positioning (PPP) by improving the positioning accuracy and reducing the convergence time because more satellites will be available. This paper discusses the performance of multi-constellation kinematic PPP based on a multi-constellation kinematic PPP model, Kalman filter and stochastic models. The experimental dataset was collected from the receivers on a vehicle and processed using self-developed software. A comparison of the multi-constellation kinematic PPP and real-time kinematic (RTK) results revealed that the availability, positioning accuracy and convergence performance of the multi-constellation kinematic PPP were all better than those of both global positioning system (GPS)-based PPP and dual-constellation PPP. Multi-constellation kinematic PPP can provide a positioning service with centimetre-level accuracy for dynamic users.

A new method of real-time cycle-slip detection and repair for BeiDou navigation satellite system (BDS) triple-frequency undifferenced observations is proposed. The method, which is based on code–phase and geometry-free phase combinations, composes linearly independent combination observations to uniquely determine the cycle-slip of original frequency phase observations in two independent steps. First, two extra-wide lane (EWL) code–phase combinations are utilised to determine the EWL combination cycle-slips. The sum of the carrier phase combination coefficients is zero. Second, the geometry-free phase combination with a non-zero sum of combination coefficients composes linearly independent combination observations with two code–phase combinations. Ionospheric delay variance systematic deviation is corrected with several previous carrier-phase observations without cycle-slip, which seriously affect the accuracy of the geometry-free phase combination observations. The BDS triple-frequency carrier-phase and code observations of different satellite types at 30 s sampling interval, which involve increased active ionospheric conditions in 1 day, are utilised to test the method. Results show that the proposed method can efficiently detect and accurately repair any cycle-slip combinations in real-time for different satellite types under active ionospheric conditions. No erroneous and leakage judgments are observed in the test.

As one of the atmosphere propagation delays, the tropospheric delay is a significant error source that should be properly handled in high-precision global navigation satellite system (GNSS) applications. We propose an improved zenith tropospheric delay (ZTD) modeling method whereby the piecewise model of the atmospheric refractivity is introduced. Compared with using the exponential model to fit ZTD in vertical direction, the ZTD piecewise model has a better performance. Based on ERA5 2.5° × 2.5° reanalysis data produced by the European Centre for Medium-Range Weather Forecasting (ECMWF) from 2013 to 2017, we establish the regional gridded ZTD model (RGZTD) using a trigonometric function for China and the surrounding areas, which ranges from 70° E to 135° E in longitude and from 15° N to 55° N in latitude. To verify the performance of RGZTD model, the ERA5 ZTD data in 2017–2018, the radiosonde ZTD data from 86 radiosonde stations over China in 2017–2018, and the tropospheric delay products on 251 GNSS stations from Crustal Movement Observation Network of China (CMONOC) in 2016–2017 are used as external compliance check data. The results show that the overall accuracy of RGZTD model is better than that of exponential model, UNB3m model, and GPT3 model. Moreover, the accuracy can be improved by about 13.4%, 7.1%, and 6.2% when ERA5 reanalysis data, radiosonde data, and CMONOC data are used as reference values, respectively. High-accuracy ZTD data can be provided because the RGZTD model takes into account the vertical variation of ZTD through the new piecewise model.