An Empirical Investigation into the Critically Low Practical Availability of Dual-Frequency Smartphone GNSS PPK

Jong Young Cho, In Ho Lee

Peer Reviewed

Abstract:	The mass-market proliferation of dual-frequency (L1/L5) GNSS chipsets in consumer-grade smartphones has sparked a paradigm shift in high-precision positioning research, promising to ”democratize” access to centimeter-level accuracy previously reserved for geodetic-grade equipment. While the inclusion of L5 carrier-phase observations theoretically enables robust Post-Processed Kinematic (PPK) performance by forming ionosphere-free combinations, this study argues that such optimism overlooks severe practical limitations inherent to smartphone hardware architecture. We present a rigorous empirical investigation into the practical availability of fixed-ambiguity solutions, revealing a critical gap between theoretical potential and real-world utility. To isolate smartphone-specific error sources from atmospheric and orbital effects, a static short-baseline (2m) experiment was conducted using a dual-band smartphone and a geodetic-grade Trimble R8-4 receiver. Data were logged for 3,600 epochs at a 1Hz interval. Processing was performed using a standard double-difference (DD) algorithm in RTKLIB (v2.4.3 b34) without smartphone-specific tuning, employing a ”Continuous” ambiguity resolution strategy and a strict Ratio-Test threshold of 3.0. Our results reveal a ”stark duality” in performance. In brief intervals where integer ambiguities were successfully fixed, the smartphone achieved a horizontal RMSE of 0.08–0.09 m, verifying its high-precision potential. However, this precision was largely inaccessible, with a global fix availability of only 9.2%. We identify a compounding failure cascade: low-gain antennas and frequent L5 signal dropouts destabilize the DD network, while aggressive power-saving duty cycling introduces non-Gaussian phase discontinuities. These artifacts ”contaminate” the Kalman filter’s covariance matrix, leading to overconfidence in erroneous float solutions and making the Ratio-Test statistically incapable of distinguishing correct integer candidates. This research demonstrates that traditional geodetic processing paradigms are fundamentally mismatched with smartphone GNSS data. We conclude by calling for a new evaluation paradigm that prioritizes fix availability over peak accuracy to realize true mobile precision.
Published in:	Proceedings of the ION 2026 Pacific PNT Meeting April 13 - 16, 2026 Hilton Waikiki Beach Honolulu, Hawaii
Pages:	338 - 349
Cite this article:	Cho, Jong Young, Lee, In Ho, "An Empirical Investigation into the Critically Low Practical Availability of Dual-Frequency Smartphone GNSS PPK," Proceedings of the ION 2026 Pacific PNT Meeting, Honolulu, Hawaii, April 2026, pp. 338-349. https://doi.org/10.33012/2026.20589
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In