Remarks on Tracking 5G NTN Signals with a GNSS-Centric SDR: CRLB, GPU/CPU-Based Correlation and C/N0-Estimation

Florian Binder, Christian A. Lichtenberger, Thomas Pany, Ottavio M. Picchi, Jón Winkel, Marco Rotoloni, Francesco Menzione, Francis Soualle

Peer Reviewed

Abstract:	This paper investigates the feasibility of acquiring and tracking 5G New Radio (5G NR) signals for Positioning, Navigation, and Timing (PNT) using a Global Navigation Satellite System (GNSS)-centric software-defined radio (SDR) receiver. A 5G NR reference waveform compliant with 3GPP specifications is integrated into the Multi-Sensor Navigation Analysis Tool (MuSNAT), enabling tracking of wideband Frequency Range 1 (FR1) signals with long coherent integration intervals. Particular emphasis is placed on the use of graphics processing unit (GPU)-based correlation to reduce processing latency while maintaining numerical compatibility with the established CPU pipeline. To ensure a meaningful performance assessment, theoretical limits are derived using the Cramer–Rao Lower Bound (CRLB). ´ The analysis builds on the generalized early–late tracking theory by Betz and Kolodziejski (2009) and incorporates extensions for block-processed and duty-cycled reference signals based on the work conducted by del Peral-Rosado et al. (2012) and Borio et al. (2011). These bounds are used as benchmarks to evaluate frequency- and code-tracking performance under simulated 5G NR Non-Terrestrial Network (NTN) conditions. A dedicated 5G NR reference waveform featuring six consecutive Positioning Reference Signal (PRS) symbols, 92.16 MHz bandwidth, and 30 kHz subcarrier spacing is generated using the Qascom QA707 simulator. Tests are performed under FR1 conditions using a specific white Gaussian noise profile. The tracking results show that graphics processing unit (GPU) and central processing unit (CPU) correlators produce equivalent punctual outputs and C/N0 estimates. Moreover, the achieved frequency- and code-tracking accuracy closely approaches the theoretical CRLB, independent of the use of the block-processing discriminator. The findings confirm the suitability of 5G NR reference signals as a high-quality PNT source and demonstrate that real-time tracking with a GNSS-centric SDR becomes feasible when GPU-based correlation is employed. Future work will address more realistic NTN propagation conditions, phase-coherence effects across full 5G frames, and validation in FR2 bands.
Published in:	Proceedings of the 2026 International Technical Meeting of The Institute of Navigation January 26 - 29, 2026 Hyatt Regency Orange County Anaheim, California
Pages:	60 - 75
Cite this article:	Binder, Florian, Lichtenberger, Christian A., Pany, Thomas, Picchi, Ottavio M., Winkel, Jón, Rotoloni, Marco, Menzione, Francesco, Soualle, Francis, "Remarks on Tracking 5G NTN Signals with a GNSS-Centric SDR: CRLB, GPU/CPU-Based Correlation and C/N0-Estimation," Proceedings of the 2026 International Technical Meeting of The Institute of Navigation, Anaheim, California, January 2026, pp. 60-75. https://doi.org/10.33012/2026.20526
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In