NLOS signal detection and correction for smartphone using convolutional neural network and variational mode decomposition in urban environment

Liu, Qi; Gao, Chengfa; Shang, Rui; Peng, Zihan; Zhang, Ruicheng; Gan, Lu; Gao, Wang

doi:10.1007/s10291-022-01369-2

Cited by 13 publications

(4 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The model first uses CNN to identify the NLOS signal. Then, the variational mode decomposition method is used to decompose the detected NLOS signals and eliminate the NLOS part [ 34 ]. Based on image characteristics, a CNN model is proposed to estimate images with synthetic multipath information generated by a GNSS signal-propagation simulator [ 35 ].…”

Section: Related Workmentioning

confidence: 99%

An Efficient Convolutional Denoising Autoencoder-Based BDS NLOS Detection Method in Urban Forest Environments

Qin,

Li,

Xie

et al. 2024

Sensors

View full text Add to dashboard Cite

The BeiDou Navigation Satellite System (BDS) provides real-time absolute location services to users around the world and plays a key role in the rapidly evolving field of autonomous driving. In complex urban environments, the positioning accuracy of BDS often suffers from large deviations due to non-line-of-sight (NLOS) signals. Deep learning (DL) methods have shown strong capabilities in detecting complex and variable NLOS signals. However, these methods still suffer from the following limitations. On the one hand, supervised learning methods require labeled samples for learning, which inevitably encounters the bottleneck of difficulty in constructing databases with a large number of labels. On the other hand, the collected data tend to have varying degrees of noise, leading to low accuracy and poor generalization performance of the detection model, especially when the environment around the receiver changes. In this article, we propose a novel deep neural architecture named convolutional denoising autoencoder network (CDAENet) to detect NLOS in urban forest environments. Specifically, we first design a denoising autoencoder based on unsupervised DL to reduce the long time series signal dimension and extract the deep features of the data. Meanwhile, denoising autoencoders improve the model’s robustness in identifying noisy data by introducing a certain amount of noise into the input data. Then, an MLP algorithm is used to identify the non-linearity of the BDS signal. Finally, the performance of the proposed CDAENet model is validated on a real urban forest dataset. The experimental results show that the satellite detection accuracy of our proposed algorithm is more than 95%, which is about an 8% improvement over existing machine-learning-based methods and about 3% improvement over deep-learning-based approaches.

show abstract

Section: Related Workmentioning

confidence: 99%

An Efficient Convolutional Denoising Autoencoder-Based BDS NLOS Detection Method in Urban Forest Environments

Qin,

Li,

Xie

et al. 2024

Sensors

View full text Add to dashboard Cite

show abstract

“…As for the multipath effect, an effective approach is based on the antenna design techniques [40], such as choke rings and dual-polarized antennas. Traditional signal processing methods that improve the structure of receiver correlators can also partially reduce multipath effects [41]. Despite many studies focused on correcting these systematic errors, some of these errors remain unmodeled.…”

Section: Introductionmentioning

confidence: 99%

Improving the ambiguity resolution with the consideration of unmodeled errors in GNSS medium and long baselines

Wang,

2024

Meas. Sci. Technol.

View full text Add to dashboard Cite

Ambiguity resolution (AR) is fundamental to achieve high-precision solution in GNSS (Global Navigation Satellite System) relative positioning. Extensive research has shown that systematic errors are associated with the performance of AR. However, due to the physical complexity, some systematic errors would inevitably remain in the observation equations even after processed with some popular models and parameterization. In the medium and long baselines, these unmodeled errors are the leading cause of the slow or even incorrect fixation of ambiguity. Therefore, to improve the AR performance in the medium and long baselines, we present a procedure with the careful consideration of unmodeled errors. At first, we develop a method to estimate the unmodeled errors based on the float ambiguity bias. Then, the overall procedure and key steps to fix the float solutions corrected by the unmodeled error estimate is designed. Finally, some real-measured baselines (from 68 km to 120 km) are utilized to validate the proposed procedure. The experimental results are analyzed and discussed from the aspects of AR and positioning, respectively. For the AR performance, the time required for the first fixation have been reduced by about 41.58% to 83.51%, from 12 to 100 min. Besides, 12.72% to 48.59% and 2.96% to 36.28% improvements of the ambiguity-fixed rate and the ambiguity-correct rate can be respectively obtained in the four baselines. As for the positioning performance, the mean values and RMSEs have improved by 0.2 to 4.8 cm (1.63% to 22.43%) and 0.2 to 2.8 cm (1.47% to 10.57%), respectively.

show abstract

“…As a result, the overall positioning accuracies have been improved. Liu et al [ 33 ] proposed a method to detect and correct the non-line-of-sight (NLOS) signals, which is an important issue in urban environments. This method is based on a convolutional neural network constructed using the original observations of smartphones providing the detection accuracy of more than 95%.…”

Section: Introductionmentioning

confidence: 99%

GNSS Observation Generation from Smartphone Android Location API: Performance of Existing Apps, Issues and Improvement

Zangeneh-Nejad

Yang

Gao

2023

Sensors

View full text Add to dashboard Cite

Precise position information available from smartphones can play an important role in developing new location-based service (LBS) applications. Starting from 2016, and after the release of Nougat version (Version 7) by Google, developers have had access to the GNSS raw measurements through the new application programming interface (API), namely android.location (API level 24). However, the new API does not provide the typical GNSS observations directly (e.g., pseudorange, carrier-phase and Doppler observations) which have to be generated by the users themselves. Although several Apps have been developed for the GNSS observations generation, various data analyses indicate quality concerns, from biases to observation inconsistency in the generated GNSS observations output from those Apps. The quality concerns would subsequently affect GNSS data processing such as cycle slip detection, code smoothing and ultimately positioning performance. In this study, we first investigate algorithms for GNSS observations generation from the android.location API output. We then evaluate the performances of two widely used Apps (Geo++RINEX logger and GnssLogger Apps), as well as our newly developed one (namely UofC CSV2RINEX tool) which converts the CSV file to a Receiver INdependent Exchange (RINEX) file. Positioning performance analysis is also provided which indicates improved positioning accuracy using our newly developed tool. Future work finding out the potential reasons for the identified misbehavior in the generated GNSS observations is recommended; it will require a joint effort with the App developers.

show abstract

NLOS signal detection and correction for smartphone using convolutional neural network and variational mode decomposition in urban environment

Cited by 13 publications

References 36 publications

An Efficient Convolutional Denoising Autoencoder-Based BDS NLOS Detection Method in Urban Forest Environments

An Efficient Convolutional Denoising Autoencoder-Based BDS NLOS Detection Method in Urban Forest Environments

Improving the ambiguity resolution with the consideration of unmodeled errors in GNSS medium and long baselines

GNSS Observation Generation from Smartphone Android Location API: Performance of Existing Apps, Issues and Improvement

Contact Info

Product

Resources

About