Two residuals can reconstruct the same large value
The mathematical example uses two coprime moduli. Switch on measurement error to see why a classic CRT can become problematic under small residual errors.
Why velocity "folds over" in Doppler space
In a chirp sequence, the Doppler phase is sampled discretely over time. If the phase change between samples is too large, different real velocities can produce the same observed discrete frequency.
Important: maximum unambiguous velocity and velocity resolution are different quantities. A waveform can have fine Doppler resolution and still alias at high velocities.
CRT combines several modulo views
If the moduli are chosen suitably, there exists exactly one number within a common range that fits all residuals.
Different chirp/PRF configurations yield different folds
When the same target Doppler is measured under several configurations with different unambiguous spans, multiple folded observations result. If they agree on the same physical target, their joint consistency can unlock the extended velocity range.
physical value
first fold
second fold
same target?
joint hypothesis
physical plausibility
Classic CRT is sensitive to residual errors and mismatches
The exact theorem expects integer residuals for known moduli. For coprime moduli, exactly one CRT solution formally still exists within the common range – but a small residual error can shift it far from the true value. In real radar data, quantization, bin mismatches, different target associations and measurement noise can additionally mean that no jointly physically plausible hypothesis remains within the given tolerances.
Bin Quantization
Peak positions are mapped onto discrete Doppler bins.
Measurement Mismatch
Different frames can accidentally couple different reflection components or targets.
Parameter Deviation
Small differences in effective resolution or timing can violate the idealized congruence.
Robust CRT extends the exact principle with error tolerance
Robust CRT methods were developed to tolerate bounded errors in the residuals. They do not, however, replace the need for correct target association and clean waveform calibration.
Ideal
Very elegant, but sensitive to inconsistent residuals.
Error Tolerance
Reconstruction stays more stable under defined error conditions.
Alternative Consistency Search
Practical radar methods can generate candidates and score them by consistency, clustering or additional measured quantities.
Velocity disambiguation is a system problem
Besides CRT, there are staggered/jittered chirps, multiple PRFs, phase-compensation-based approaches, clustering and range-/track-supported methods. The best choice depends on waveform, MIMO scheme, compute budget, SNR and multi-target scenario.
Unwrapping only works if the measurements actually belong together
Multi-Target
Residuals from different targets must not be combined with each other.
Acceleration
If velocity changes between the configurations, the joint hypothesis shifts.
Low SNR
Unstable peak positions increase residual error.
Clutter
Strong background peaks can lead to false candidates.
Waveform Coupling
Range, velocity, MIMO and timing are not independent.
No Ozense Disclosure
The page explains methods, not the concrete internal waveform planning.
Frequently asked questions on Doppler ambiguity and CRT
What is unambiguous velocity?
The velocity range within which a Doppler measurement can be unambiguously interpreted without folding.
Is CRT mandatory for FMCW?
No. It is one possible approach. Many other waveform and disambiguation methods exist.
Why do you need coprime moduli?
In classic CRT, coprime moduli produce a unique joint reconstruction range with size equal to their product.
What happens with measurement errors?
The exact CRT can fail badly. Robust-CRT or search-based methods address bounded errors, but have their own assumptions and limits.
Sources
- Gonzalez, A. et al. (2021). Doppler disambiguation in MIMO FMCW radars with binary phase modulation. IET Radar, Sonar & Navigation. DOI:
10.1049/rsn2.12063. - Li, X.; Liang, H.; Xia, X.-G. (2009). A Robust Chinese Remainder Theorem with Its Applications in Frequency Estimation From Undersampled Waveforms. IEEE Transactions on Signal Processing. DOI:
10.1109/TSP.2009.2025079. - Li, X.; Xia, X.-G.; Liang, H. (2010). A Robust Chinese Remainder Theorem with Its Applications in Moving Target Doppler Estimation. IEEE Radar Conference. DOI:
10.1109/RADAR.2010.5494420. - Wang, W.-J.; Xia, X.-G. (2010). A Closed-Form Robust Chinese Remainder Theorem and Its Performance Analysis. DOI:
10.1109/TSP.2010.2066974.