A Survey of Machine Learning Techniques for Improving Global Navigation Satellite Systems

TL;DR

This paper addresses the GNSS positioning challenges posed by multipath, NLOS, and environmental variability by surveying a broad spectrum of machine learning techniques. It synthesizes supervised, unsupervised, deep learning, and hybrid approaches applied to signal analysis, anomaly detection, sensor fusion, and time-series forecasting, highlighting concrete performance gains and practical limitations. Key contributions include post-2021 developments, diverse GNSS use cases, and a critical appraisal of methodological strengths, gaps, and prospective research directions. The work informs researchers and practitioners on effective ML strategies for robust GNSS positioning and outlines actionable opportunities for future integration and standardization.

Abstract

Global Navigation Satellite Systems (GNSS)-based positioning plays a crucial role in various applications, including navigation, transportation, logistics, mapping, and emergency services. Traditional GNSS positioning methods are model-based and they utilize satellite geometry and the known properties of satellite signals. However, model-based methods have limitations in challenging environments and often lack adaptability to uncertain noise models. This paper highlights recent advances in Machine Learning (ML) and its potential to address these limitations. It covers a broad range of ML methods, including supervised learning, unsupervised learning, deep learning, and hybrid approaches. The survey provides insights into positioning applications related to GNSS such as signal analysis, anomaly detection, multi-sensor integration, prediction, and accuracy enhancement using ML. It discusses the strengths, limitations, and challenges of current ML-based approaches for GNSS positioning, providing a comprehensive overview of the field.

Figures (8)

• Figure 1: Illustration of SVM from figsvm. SVM is a supervised learning algorithm that finds the hyperplane that best separates different classes with the maximum margin. It uses support vectors and kernels to optimize the separation boundary in both linear and non-linear classification tasks.
• Figure 4: An example illustrating how KNNs are used in classification tasks figknn. KNN is a non-parametric learning algorithm that classifies new cases based on the majority vote of the $k$ most similar instances from the training data, often using distance metrics like Euclidean distance to determine similarity.
• Figure 5: Reinforcement Learning (RL) involves agents learning to make decisions by taking actions in an environment to maximize cumulative reward. Through trial and error, the agent refines its policy to achieve optimal outcomes. Figure adapted from sutton1998reinforcement.
• Figure 6: Illustration of the Transformer architecture from vaswani2017attention. While this architecture has revolutionized language models, it has been used recently to capture temporal and spatial dependencies in GNSS measurements and improve positioning accuracy.
• Figure 10: The main error sources in GNSS positioning in urban environments include Non-line-of-sight (NLOS) errors, blocked signals, and MP from reflected signals. By using ML techniques that can classify the signals into these categories, we can improve the accuracy and reliability of positioning. Figure adapted from gnss_errors.