GESTALT® is a Global Navigation Satellite Systems Signal Testbed consisting of:
A physical infrastructure for research and development of software defined GNSS receivers, featuring:
- A set of GNSS antennas mounted in a platform at the roof of the CTTC building.
- GNSS Signal generation equipment for controlled experiments.
- A set of RF front-ends, in charge of amplification, filtering, down-shifting to baseband, digitizing and interfacing with the computing platform.
- A computing platform where the software receiver is executed.
An open source project, named GNSS-SDR, that implements the GNSS software-defined receiver used in GESTALT® and provides an open development ecosystem consisting of:
- A website at http://gnss-sdr.org
- Public source code repositories with and associated version control system at https://github.com/gnss-sdr/gnss-sdr
- A public mailing list.
GESTALT® allows for rigorous, transparent, fair and replicable testing of signal processing algorithms and receiver architectures. The testbed includes hardware, software, and networking components, constituting a state-of-the-art facility for research and development of next-generation’s GNSS receivers.
The swiftly evolving landscape of GNSS signals and systems demands rapid prototyping tools in order to explore receivers’ full capability, including radically new uses of those signals. That flexibility is hard to find in today’s GNSS receiver technology, mainly driven by application-specific integrated circuits (ASIC) and system-on-chip (SoC) implementations with high development costs and very limited degree of reconfigurability, thus hampering experimentation and fair trials of new approaches. Manufacturers are incorporating new features to their commercial receivers at steady pace: most low-cost, mass-market GNSS receivers are already multi-constellation (GPS and GLONASS) but still not multiband. In contrast, professional receivers are mostly dual-band, with some triple-band model (Novatel’s OEM628) already available. In all cases, modern GNSS receivers’ performance heavily relies on assistance data from external systems (e.g., cellular and WiFi networks) in order to shorten the time-to-first-fix or enhance their navigation performance via the application of high-accuracy algorithms such as Real Time Kinematics (RTK) or Precise Point Positioning (PPP).
In spite of the indubitable interest of multi-frequency receivers, an experimentation platform at the signal processing level for dual and triple band GNSS receivers was still missing. GESTALT® is equipped with broadband, geodetic grade antennas; GNSS signal generators for controlled experiments; state-of-the-art radio-frequency front-ends able to work concurrently in three GNSS frequency bands, with configurable bandwidth, frequency downshifting and filtering; digitation working at sample rates as high as 80 Msps with 8-bit, coherent I/Q samples; high-speed interfaces to a host computer; and an open source GNSS software receiver in charge of signal processing and generation of suitable outputs in standard formats. A non-exhaustive list of suitable applications that could be deployed in such platform is: signal recording and playback, algorithm development and validation, interference monitoring, array processing, a GNSS reference station, antenna/front-end assessment, reflectometry, low-cost high-accuracy solutions, space weather monitoring, ionospheric mapping, customized guidance of unmanned vehicles or GNSS-based cloud services.
A key aspect of GESTALT® is its openness. In addition to the fact that it can be fully operated remotely, the core software receiver engine in charge of all the digital signal processing chain is an open source project with a lively community of users and developers. Accordingly, a partial testbed replication can be done on a limited budget with commodity computers and low cost, over-the-counter antennas and radio-frequency front-ends. This allows both for reproducible research and short assessment and validation times, ultimately shortening the gap between ideas for new uses of GNSS signals and user-driven, market-ready products and services.
General features
- Multiplatform (Linux and Mac OS X, 32 and 64 bit architectures, ARM processors).
- Multithreaded.
- Multiband.
- Works with files and several RF front ends, including those compatible with the Universal Hardware Driver (UHD).
- Acquisition of GPS L1 C/A, Galileo E1B and E1C signals. GPS L2C (M) and Beidou B1I are ongoing work.
- SIMD-enabled for most popular processors.
- Implemented tracking loops: DLL + PLL, DLL + PLL/FLL, VEML, with most popular discriminators.
- Connection to Matlab/Simulink via TCP for rapid prototyping and algorithm validation.
- Demodulation and decoding of the navigation message GPS NAV and Galileo INAV.
- Computation of PVT (Position – Velocity – Time) solution in real-time.
- Position solution exportable to KML files (can be opened by Google Earth and other similar tools) and to a serial port via NMEA.
- Generation of RINEX files (observables and navigation), v2.1 and v3. RTCM 3.2 is ongoing work.
- Source code released under the General Public License.