The GNSS SDR Metadata Standard is accompanied by an exemplary source code and test data sets.
A C++ source code implementing the standard and providing an application programming interface (API) for reading metadata data and reading IF samples for multiple platforms can be found at https://github.com/IonMetadataWorkingGroup/GNSS-Metadata-Standard/tree/master.
For questions or comments regarding the C++ code and the API, please contact the Working Group co-chair:
James Curran: jamestcurran(at)ieee.com
This page also contains sampled data sets of live-sky or RF constellation simulator based GNSS signals produced by various software defined radio (SDR) data collection systems. The main purpose of these data sets is to continue developing and to provide test vectors when integrating the API into own source code. Metadata files are XML based and the file extension must be xml or ending with an ‘x’. Binary sample files may use any extension. They metadata files include the considered RF bands, sampling rate, bandwidth, number of bits per sample and additional information describing the recorded data.
The data sets are organized alphabetically. The files are intentionally small to keep download times and server traffic to a minimum, but typically allow to obtain a position fix. The files are by no means fully representative of the capabilities of the recording systems; they only serve to illustrate various data formats and recording strategies. We welcome requests for additional SDR files to be hosted on this page – especially if data formats are different to those provided. We hope to extend the number of data sets available in this page over time.
For questions or comments regarding the sample data files, please contact:
Markel Arizabaleta: markel.arizabaleta(at)unibw.de
These data sets were recorded from a Spirent GSS8000 RF constellation simulator in a static scenario providing GPS and Galileo satellites on three frequency bands (L1, L2, L5). Details about the simulated position and signals can be found in the following screenshot:
Fraunhofer’s "Flexiband" USB front-end was used in its triband configuration III-1b, capturing L1/E1bc with 18 MHz bandwidth and 20MSPS@4bit I/Q sampling rate, L2/L2C with 18 MHz bandwidth and 20MSPS@4bit I/Q sampling rate and L5/E5a with 38 MHz bandwidth and 40MSPS@4bit I/Q sampling rate. Further information about the Flexiband front-end can be found in the following papers:
Additional information URL: http://www.iis.fraunhofer.de/flexiband
POC for data set: alexander.ruegamer(at)iis.fraunhofer.de
This data set was recorded also in Oegstgeest, Netherlands, with a Great Scott Gadgets HackRF device. HackRF is a SDR peripheral capable of transmission or reception of signals between 1 and 6GHz, for more information see the additional information. The recollected data includes the GPS L1 band with 10MHz bandwidth and 10MSPS@8bit I/Q sampling rate. This case is similar to the BladeRF, in the sense of detecting the GPS L1 signals (PRN 2, 5, 6, 7, 9, 13, 29, 30).
Additional information URL: https://greatscottgadgets.com/hackrf/
The data set was captured with a roof top antenna at the IFEN office building in Munich/Germany. It includes wideband (20.48 MHz sampling rate) L1/E1 and L2 signals. GPS signals are included as well as signals from four Galileo satellites (PRN 11, 12, 19, 20). IFENs software receiver SX3 was used to capture the data.
Additional information URL: http://www.ifen.com
This data sets were generated with the “SatGen” software from LabSat. The software allows to generate GPS, GLONASS, and/or Beidou signals with different quantization levels. This data collection includes the GPS L1 band with 15MSPS@1bit, 2bits, and 3bits I/Q sampling rate. In this case, the following GPS L1 signals are detected with the PRN 1, 3, 8, 11, 14, 17, 18, 19, 22, 23, 31 and 32.
Additional information URL: https://www.labsat.co.uk/index.php/en/customer-area/software-firmware
The data set was captured with the LimeSDR, an open source software-define radio (SDR) platform that supports any type of wireless communication standard such as UMTS, LTE, etc. It also can be used to detect GPS signals, which are sampled at 10MSPS@16bit I/Q sampling rate. The data was collected on Oct 15, 2017 and detects the GPS L1 signals with the PRN 2, 5, 6, 7, 9, 13, 29 and 30.
Additional information URL: https://myriadrf.org/projects/limesdr/
This data sets was recorded with NT1065_USB3 device of NTLab. It includes the GPS L1, L2 and GLONASS L1, L2 bands with 53 MSPS@2bit at IF level. The acquired signals belonging to the GPS L1 band are the PRN 4, 8, 10, 11, 18, 27 and 32.
Additional information URL: http://ntlab.com/index.htm
This data set was recorded from a BladeRF SDR of Nuand. It includes the GPS L1 band with 8MHz bandwidth and 10MSPS@16bit I/Q sampling rate. The data was collected on Sep 12, 2017 in Oegstgeest, Netherlands. In this case the following GPS L1 signals are detected with the PRN 30, 7, 5, 13, 15, 2, and 27.
Additional information URL: https://nuand.com/
This data set exemplifies the temporally split recording concept described in the standard. Four IF-sampled data streams are multiplexed into a single lane. Each sample is stored at 1-bit quantization. Each RF band is coherently sampled at 56.32 MSPS. The data set contains several GPS and Galileo signals collected on Feb 26, 2013 in Athens Ohio using a NovAtel GPS-703-GGG triple-frequency antenna. Further information about the Ohio University TRIGR can be found in the following paper: Gunawardena, Sanjeev, van Graas, Frank, "Multi-Channel Wideband GPS Anomalous Event Monitor," Proceedings of the 24th International Technical Meeting of The Satellite Division of the Institute of Navigation (ION GNSS 2011), Portland, OR, September 2011, pp. 1957-1968.
Additional information URL: http://ChameleonChips.com
POC for data set: Sanjeev Gunawardena: sanjeev(at)chameleonchips.com
Data compliant with the current version of the C++ code (no temporal splitting implemented):
Test data illustrating temporal splitting supported in a future release of the C++ code:
The data set recorded with a RTL_SDR. This collection includes the GPS L1 band with 2.048MHz bandwidth and 2.048MSPS@8bit I/Q sampling rate. The data was collected on Sep 10, 2017 also in Oegstgeest, Netherlands. The acquired signals belong to the GPS L1 band (PRN 5, 13, 15, 20, 21, 28, 30).
Additional information URL: https://greatscottgadgets.com/hackrf/
SiGe radio front end
This data set was recorded with a SiGe GN3S Sampler v3, a research device co-developed by SiGe and the GNSS Lab at the University of Colorado. This data collection includes the GPS L1 band with 16.368MSPS@8bit sampling rate. The samples were collected on May 23, 2013. The acquired signals belong to the GPS L1 band (PRN 1, 7, 8, 9, 11, 17, 28).
These data collections were recorded in Shanghai, China, on Aug 28, 2017. Two different collections are given below, in one of them the Galileo E1 and GPS L1 signals (PRN 2, 5, 6, 7, 13, 19, 29) are collected and in the other one the data from Beidou2 B1 band is collected. These data sets were sampled for 60 seconds, with a sampling rate of 25MSPS@4bit at IF level and a bandwidth of 8MHz for the Galileo and GPS signals and 8MHz for the Beidou2 signals.
POC for data set: xin.chen(at)sjtu.edu.cn
This data set was recorded with a Universal Software Radio Peripheral (USRP) 2950 from National Instruments. This data collection includes the GPS L1 band with 100MSPS@16bit I/Q sampling rate. In this file only the signal of the satellite with PRN 30 will be available. For the collection of the data a 2.4m dish antenna was used.
Additional information URL: http://www.ni.com/de-de/support/model.usrp-2950.html
The IF sample files have been processed by various SDRs and the resulting RINEX observations files are available for download at the page linked below. These files serve as reference to better understand the signal content within the IF sample files and to encourage comparisons between various SDRs. Interested research groups are invited to post their own processing results.