High-precision GPS vehicle tracking to improve safety.

Humphreys, Todd E.; Murrian, Matthew J.; Gonzalez, Collin W.; Pesnya Jr., Kenneth M.; Shepard, Daniel P.; Kerns, Andrew J. · 2016 · ROSA P / University of Texas at Austin. Data-Supported Transportation Operations & Planning Center (D-STOP)

archive: archived pipeline: cataloged verified

Get this paper ↗ (full text — opens at the source; we link to it, we don't host it)

Summary

This paper addresses the challenge of enabling low-cost, centimeter-level Global Navigation Satellite System (GNSS) positioning for mass-market applications, particularly automated vehicles and safety-critical systems. While carrier-phase differential GNSS (CDGNSS) has long provided high precision for surveying and agriculture, its high cost and slow convergence times have hindered adoption in consumer markets. The authors argue that achieving rapid ambiguity resolution with inexpensive, single-frequency receivers requires a dense reference network to minimize ionospheric uncertainty, which is the primary bottleneck for fast positioning in challenging urban environments. To investigate this, the researchers analyzed data from a hypothetical network of 23 high-quality reference stations in California to model the relationship between network density and ionospheric error uncertainty. They determined that an average inter-station spacing of 20 km is required to reduce network-side errors below the level of rover receiver multipath. Based on these findings, the team deployed the "Longhorn Reference Network" in Austin, Texas, consisting of low-cost, solar-powered, dual-frequency reference stations with tighter spacing than the theoretical minimum to provide redundancy. They also developed a low-cost rover system using software-defined radios and dual antennas to test vehicle positioning and heading. The study demonstrated a live vehicle lane-departure warning system along a 1-mile route near the University of Texas campus, which included both open-sky and signal-blocked areas. The system achieved centimeter-accurate positioning with high reliability. In conditions with 16 visible satellites, 96% of measurement epochs yielded correct, internally validated solutions when using a dual-antenna configuration. Even in worst-case scenarios with only 12 visible signals, the system maintained robust heading accuracy (94% valid solutions), though position validation rates dropped. The results showed that the dense reference network effectively suppressed atmospheric errors, allowing for instantaneous, sub-decimeter positioning without the need for long convergence periods or expensive multi-frequency hardware. The significance of this work lies in demonstrating that dense reference networks can enable mass-market precise positioning by compensating for the limitations of low-cost user equipment. The findings suggest that densifying permanent reference networks, particularly in urban areas, is a viable strategy to support automated driving and safety applications. By reducing ionospheric uncertainty to under 2 millimeters, such networks allow for rapid integer ambiguity resolution, making high-precision GPS accessible for real-time safety monitoring, such as detecting driver distraction or drowsiness, and enabling reliable lane-keeping for autonomous vehicles.

Key finding

A dense GNSS reference network with approximately 20-kilometer station spacing enables centimeter-accurate vehicle positioning with over 92 percent reliability in light urban environments using low-cost single-frequency receivers.

Methodology

field_study

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed. Discovered via bulk_ingest_rosap on 2026-05-23 (6 acquisition events logged).

Summary generated by qwen3.6-27b-prismaquant on 2026-06-10; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• lane positioning