FMCW RADAR
High-resolution ranging and velocity measurement using frequency-modulated continuous-wave techniques.
A Ka-band active electronically scanned array radar platform being developed for real-time detection, tracking, and beam-steered sensing.
High-resolution ranging and velocity measurement using frequency-modulated continuous-wave techniques.
Electronically controlled transmit and receive channels designed for rapid beam steering and multi-target sensing.
Range-Doppler processing, target detection, and tracking on embedded compute hardware.
FARTTS integrates waveform generation, Ka-band frequency conversion, an electronically steered array, SDR acquisition, and real-time processing into a single experimental radar platform.
The prototype is under active development. Current work spans RF architecture, custom antenna arrays, reference distribution, signal processing, hardware integration, and test planning.
Support for the project’s RF conversion and signal-chain development.
Support related to the software-defined radio portion of the system.
Full-wave electromagnetic simulation support for antenna and array development.
RF and microwave component support used in the radar signal chain.
Test and measurement support for future RF characterization and validation.
Funding helps cover remaining beamforming hardware, RF interconnects, LO hardware, antenna fabrication, and other integration costs.
FARTTS is a Ka-band FMCW radar platform centered on electronically steered transmit and receive apertures, coherent timing, SDR acquisition, and embedded signal processing.
This is the current planned hardware architecture for the FARTTS prototype. Hover over a commercially identified component to see its name, then click it to open the manufacturer’s official product page.
Dedicated upconversion and downconversion stages translate between the SDR/IF domain and the Ka-band RF operating region. The architecture is designed around a stable shared reference and controlled LO generation.
Multiple transmit and receive channels drive a phased-array architecture. Element spacing, channel phase control, feed-network layout, and calibration directly shape the beam.
Complex I/Q data is captured in the digital domain for waveform analysis, beat-frequency extraction, range-Doppler processing, and algorithm development.
The processing pipeline combines range, velocity, and angle information to form persistent target tracks and a real-time operator display.
Current FARTTS frequency planning, candidate FMCW operating modes, Doppler estimates, and USRP B210 bandwidth budgeting. These values are design calculations and will be refined as bench characterization and integrated testing progress.
The current plan uses a 1.0 GHz intermediate frequency and the ×4 LO path inside the ADMV1013 / ADMV1014 conversion chain. Sweeping the external LO from 6.625 to 6.875 GHz produces the planned 27.5 to 28.5 GHz RF sweep.
With fIF = 1.0 GHz, the lower RF edge uses 6.625 GHz LO and the upper RF edge uses 6.875 GHz LO.
A 1.0 GHz FMCW sweep gives the same theoretical range resolution across the candidate modes; changing chirp time changes slope and beat-frequency scaling.
All three candidate modes retain the same 1.0 GHz sweep bandwidth. Increasing chirp duration reduces the FMCW slope, which reduces beat frequency for a target at the same range and makes longer-range returns easier to keep inside the SDR capture bandwidth.
| Mode | Chirp time | Sweep slope | 100 ft range beat | 1000 ft range beat | Idealized bandwidth-limited range* |
|---|---|---|---|---|---|
| Fast / Short Range | 50 µs | 2.0 × 1013 Hz/s 20 THz/s |
≈ 4.06 MHz | ≈ 40.6 MHz | ≈ 420 m ≈ 1,376 ft |
| Medium | 500 µs | 2.0 × 1012 Hz/s 2 THz/s |
≈ 406 kHz | ≈ 4.06 MHz | ≈ 4.2 km ≈ 2.6 mi |
| Slow / Long Range | 1 ms | 1.0 × 1012 Hz/s 1 THz/s |
≈ 203 kHz | ≈ 2.03 MHz | ≈ 8.4 km ≈ 5.2 mi |
The estimate reserves roughly 42.6 kHz for a 500 mph radial Doppler shift and then uses the nominal 56 MHz B210 real-time bandwidth. Final usable range will depend on actual spectral placement, sample rate, filter rolloff, guard bands, SNR, target RCS, antenna gain, receiver noise figure, processing gain, and the measured analog front end.
For a stationary target, the approximate FMCW range beat is fR = 2SR/c, where S is chirp slope. The table shows why the long-range modes use slower sweeps.
| Target range | 50 µs fast chirp | 500 µs medium chirp | 1 ms slow chirp | Nominal 56 MHz B210 comparison |
|---|---|---|---|---|
| 100 ft / 30.48 m | ≈ 4.06 MHz | ≈ 406 kHz | ≈ 203 kHz | Comfortable |
| 1000 ft / 304.8 m | ≈ 40.6 MHz | ≈ 4.06 MHz | ≈ 2.03 MHz | Mode dependent |
| 1 mile / 1.609 km | ≈ 214.5 MHz | ≈ 21.45 MHz | ≈ 10.73 MHz | Use slower chirp |
| 2 miles / 3.219 km | ≈ 429.2 MHz | ≈ 42.9 MHz | ≈ 21.46 MHz | Medium / slow candidate |
| 5 miles / 8.047 km | ≈ 1.07 GHz | ≈ 107.3 MHz | ≈ 53.6 MHz | Slow mode only; very little nominal margin |
The B210 specification is up to 56 MHz of instantaneous real-time RF bandwidth at up to 61.44 MS/s quadrature. That does not automatically mean every configuration provides a full 56 MHz of usable one-sided beat-frequency offset. The 1.0 GHz IF center, digital tuning, sample rate, analog/digital filters, and guard bands will be selected so the expected beat spectrum stays inside the actual measured passband.
At Ka-band, radial motion adds a Doppler term approximately fD = 2v/λ. Across 27.5–28.5 GHz, the wavelength is roughly 10.9–10.5 mm.
| Radial speed | Approx. speed | Doppler shift across 27.5–28.5 GHz |
|---|---|---|
| 20 mph | 8.94 m/s | ≈ 1.64–1.70 kHz |
| 100 mph | 44.7 m/s | ≈ 8.2–8.5 kHz |
| 500 mph | 223.5 m/s | ≈ 41.0–42.6 kHz |
The USRP B210 provides continuous 70 MHz–6 GHz RF coverage and up to 56 MHz of real-time bandwidth. The planned 1.0 GHz IF is comfortably inside its tuning range. Actual usable beat span will be validated with the selected sample rate and filter settings.
A fast 50 µs chirp produces larger beat frequencies and a high update rate for short-range tracking. Extending the chirp to 500 µs or 1 ms reduces slope and compresses long-range beat frequencies into the available SDR bandwidth.
At roughly 0.25 m separation, a simple free-space calculation gives about 49.1–49.5 dB of path loss across 27.5–28.5 GHz. For closely spaced phased arrays, however, actual coupling can differ substantially because of aperture pattern, polarization, mounting, mutual coupling, reflections, and near-field behavior. Measured TX-to-RX isolation or simulated coupling is more useful than FSPL alone.
Digital self-interference cancellation is planned to suppress residual TX leakage after digitization. It cannot undo front-end compression or ADC clipping, so passive isolation, converter gain/attenuation, receiver linearity, and available dynamic range still matter before DSIC is applied.
Simple FSPL is not the final model for target detectability. A reflected radar return includes two-way propagation plus target radar cross section, antenna gains, transmit power, receiver noise, losses, and processing gain. That link budget will be refined separately from the direct-leakage calculation.
Once the RF front end and arrays are integrated, FARTTS will measure direct-path isolation and the usable B210 passband, then use those measurements to set practical chirp limits, gain states, guard bands, and DSIC requirements.
A running view of the engineering work behind FARTTS: antenna design, RF integration, timing, signal processing, bench validation, and eventual field testing.
Define the transmit/receive conversion chain, frequency plan, LO strategy, timing/reference architecture, and channel count.
Tune patch geometry, impedance, bandwidth, substrate, and feed structure through full-wave EM simulation.
Scale the optimized element into arrays while controlling spacing, coupling, routing, phase length, and fabrication constraints.
Validate RF acquisition, shared frequency/time references, and stable I/Q capture at the system level.
Implement FMCW beat processing, FFT stages, target extraction, clutter mitigation, and track formation.
Perform bench characterization followed by controlled outdoor detection and tracking tests as the hardware matures.
FARTTS is being developed by a multidisciplinary electrical engineering team working across system architecture, RF, antennas, software, digital signal processing, power systems, SDR development, integration, and technical documentation.
Systems Architecture · RF · Antenna Design · Software
Leads the overall system design and integration, antenna development, RF architecture, evaluation-board control software and CSV configuration, and initial USRP / GNU Radio development.
jason@fartts.comDigital Signal Processing · Software-Defined Radio
Focused on DSP development and expanding the project’s SDR-based signal acquisition and processing work.
william@fartts.comPower Systems · GNU Radio Algorithms
Works primarily in power systems while also contributing algorithms intended for implementation inside custom GNU Radio processing blocks.
diego@fartts.comIntegration · Hardware Bring-Up · Test Support
Supports subsystem integration and cross-functional hardware testing as the prototype moves from individual components toward a complete radar system.
rex@fartts.comDocumentation · Reports · Technical Communication
Focused on the writing and documentation side of the project, including technical reports, project documentation, and written deliverables.
john@fartts.comFARTTS has received hardware, software, and test-resource support from organizations that have helped move the prototype forward. The items below are listed specifically so the contribution is clear.
These entries use the exact quantities and support descriptions currently provided by the team. Product links go to the manufacturer’s official page when a specific product is identified.
Software-defined radio platform.
Evaluation hardware contributed to the RF signal chain.
Evaluation hardware contributed to the RF signal chain.
90° hybrid couplers.
One year of Ansys HFSS electromagnetic simulation access.
Future test-equipment support.
Recognition is intended to reflect meaningful support without creating a rigid public dollar threshold. Depending on the contribution, supporters may be acknowledged through the project website, project posters and presentations, demonstrations, technical reports, application notes, GitLab/project documentation, and subsystem-specific hardware or resource acknowledgments.
Meaningful supporters can receive a dedicated listing with the organization name, what they contributed, and an official product or company link when appropriate.
Supporters may be acknowledged in public-facing project posters, presentations, demonstrations, reports, and other technical material.
When hardware, software, fabrication, test equipment, or another resource directly enables a subsystem, the contribution can be credited alongside that part of the project.
Support can take the form of funding, RF hardware, fabrication, interconnects, test-equipment access, technical review, or other engineering resources.
Questions about FARTTS, technical collaboration, sponsorship, or project support can be sent directly to the project point of contact.
This static version opens your default email application. A server-backed contact form can be connected after hosting is selected.