Topic: Student Papers from ION GNSS+ 2025
Speaker 1: Paul Dobre
Title: Machine Learning Model Uncertainty in GNSS Positioning
Abstract: GNSS positioning methods such as Kalman filters, factor graph optimization, and weighted least squares (WLS) have recently been complemented by machine learning (ML) models aimed at improving positioning accuracy and robustness. ML has been applied to GNSS for signal classification, anomaly detection, environmental inference, and position correction. However, in challenging conditions—such as urban canyons or unfamiliar scenarios—ML models face epistemic (model) and aleatoric(data) uncertainty, which can result in overconfident yet incorrect predictions that compromise system integrity. This work proposes an uncertainty-aware ML framework to enhance GNSS positioning error estimation by quantifying and incorporating both epistemic and aleatoric uncertainty into the model output. A spatial transformer is used to predict GNSS positioning error based on satellite-specific observation features. The model output includes both a positioning error estimate and its associated uncertainty, which enables more reliable integration with traditional GNSS solutions. Uncertainty is quantified through a model ensemble approach that aggregates predictions from multiple models to estimate uncertainty. The benefits of incorporating uncertainty include better anomaly handling, increased model interpretability, improved retraining strategies via active learning, and the ability to fall back to traditional methods in high-uncertainty situations. The preliminary experimental results demonstrate that incorporating uncertainty improves positioning interpretability and robustness compared to a standard ML enhanced GNSS pipeline through effectively identifying out-of-distribution scenarios and high input noise and guiding system fallback decisions
Bio: Paul is currently a M.Sc. student at Intelligent Navigation and Mapping Lab at University of Calgary. His main research interests are machine learning in GNSS and autonomous driving
Speaker 2: Shichuang Nie
Title: Mamba Based GNSS Cycle Slip Detection for the Single Frequency Receiver
Abstract: Cycle slips remain a dominant error source in carrier-phase Global Navigation Satellite System (GNSS) positioning and can severely impair accuracy if they are not detected and repaired. Although robust algorithms exist for dual-frequency receivers, low-cost single-frequency units still rely primarily on Doppler-aided techniques, whose performance declines at low sampling rates (e.g., 1 Hz) and whose quality-control mechanisms are rudimentary. This study investigates whether Mamba—a recent sequence-modeling architecture that offers transformer-level contextual capacity with linear complexity—can ameliorate these limitations. We train Mamba to (i) assess the validity of Doppler-aided cycle-slip detections and (ii) estimate potential repair integer. Tests on real GPS L1 data show that the model markedly improves the reliability assessment of Doppler-derived decisions; however, its slip-correction predictions are limited, owing to the weak statistical connection between the input features and the integer nature of cycle slips. These findings highlight both the promise of modern deep-learning models for quality monitoring in low-cost GNSS and the need for richer feature representations to achieve complete cycle-slip correction.
Bio: Shichuang is currently a Ph.D. student at Intelligent Navigation and Mapping Lab at University of Calgary. He obtained his bachelor’s degree at Wuhan University in 2023. His main research interests are deep learning in GNSS, multi-sensor integration and autonomous driving.
Location: Room 207 – Engineering Block G (ENG), University of Calgary Campus
Date: Thursday November 28, 2024
Time: Doors will open at 11:45am, presentation beginning shortly after noon
Cost: $20 non-members, $18 members, $15 grad students, undergrad students $10, includes a light lunch and refreshments. All proceeds go towards two annual scholarships for students attending the University of Calgary