CoChat

Distribution-Aware Adaptive Ambiguity Resolution for Smartphone RTK in Mixed GNSS Environments

Vishwanath, Ananya · 2026 · Schulich School of Engineering
Cited by 0

The DOI matches its DataCite record; 1 of the 1 references it cites were found in a registry or an index.

Abstract

This thesis investigates observation-level inconsistencies and ambiguity-resolution chal-lenges in smartphone-based GNSS positioning. In Android devices, pseudorange genera-tion depends on how receiver clock parameters are handled across logging and conversion tools, where differences in FullBiasNanos (FBN) treatment introduce systematic timing offsets that propagate into positioning solutions. In addition, variations in satellite visibility, multipath, and signal strength across environments lead to changes in ambiguity-residual behavior that are not captured by conventional ambiguity resolution approaches. To address these limitations, this thesis first analyzes four commonly used GNSS log-ging and conversion tools: Google GNSSLogger, Geo++ RINEX Logger, CSV2RINEX, and raw_android_to_rinex. The evaluation is conducted using controlled open-sky datasets collected with multiple Google Pixel devices, including the Pixel 4, Pixel 5, and Pixel 7 Pro. Reconstructed pseudorange comparisons show that distinct and often undocu-mented FBN strategies—(1) fixed versus epoch-varying bias application, (2) truncation of nanosecond-level digits, and (3) floating-point precision loss—introduce systematic offsets of approximately 6–150 m. Single Point Positioning (SPP) results indicate that these offsets map predominantly into the estimated receiver clock term, while position components remain stable at the millimeter-to-centimeter level under consistent geom-etry. Temperature-controlled experiments further show a thermal dependence of the smartphone clock, with the FBN drift rate decreasing as device temperature increases. An adaptive Best Integer Equivariant (BIE) framework is then developed for mixed-environment Real-Time Kinematic (RTK) processing. A GNSS dataset spanning highway, suburban, and downtown conditions is constructed using visibility, Sinal-to-Noise Ratio (SNR), and geometry-based features, and a lightweight LSTM classifier provides per-epoch environment labels with 97.6–99.1% training accuracy and 89% accuracy on an independent suburban–highway test trajectory. Standardized ambiguity residuals are modeled using Gaussian, contaminated-normal, and Student-𝑡 distributions to define environment-specific weighting within the adaptive, distribution-aware BIE framework. Evaluation on suburban–highway trajectories collected with a Google Pixel 7 Pro shows that the adaptive BIE solution limits larger variations observed in float and fixed processing using Least-squares AMBiguity Decorrelation Adjustment (LAMBDA), improving 2D and 3D Root Mean Square Error (RMSE) by up to 14.5% and 10.3%, respectively. Together, these results provide a systematic characterization of smartphone pseudorange formation and demonstrate that environment-aware ambiguity modeling improves RTK reliability across mixed environments.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

DataCite · 10.11575/prism/51260

Cites

1 of 1 references verified0 citation errors · 0 not checkable

Statistics

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Metadata

Authors
Vishwanath, A.
Venue
Schulich School of Engineering
Publisher
Schulich School of Engineering
Published
2026-04-15
References
1

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.