Ambiguity-Aware Waveform
Understanding Ambiguity-Aware Waveform Design
The Ambiguity Function |χ(τ,ν)|2 is the two-dimensional matched filter output as a function of time delay τ (proportional to range) and Doppler shift ν (proportional to radial velocity). An ideal waveform would produce a single infinitely narrow spike at the true target location with zero response elsewhere, but Woodward's theorem proves the total volume under the AF surface is constant. Suppressing sidelobes in one region necessarily redistributes energy to another.
LFM (chirp) waveforms have a ridge-shaped AF: excellent range resolution (set by bandwidth) but with a diagonal coupling slope that maps Doppler errors into range errors. Phase-coded waveforms (Barker, polyphase) can produce thumbtack-like AFs with low sidelobes but require high peak power. Ambiguity-aware design selects waveform parameters, and sometimes combines waveforms across a CPI, to achieve the best trade-off for the operational scenario.
χ(τ,ν) = ∫ u(t)·u*(t+τ)·ej2πνt dt
Range resolution:
ΔR = c / (2B)
Velocity resolution:
Δv = λ / (2TCPI)
LFM range-Doppler coupling:
ΔRerror = c·fd / (2·chirp rate)
Example: 1 kHz Doppler, 100 MHz/ms chirp: ΔR ≈ 1.5 m error
Waveform AF Characteristics
| Waveform | AF Shape | Range-Doppler Coupling | Peak Sidelobe | Best For |
|---|---|---|---|---|
| Simple pulse | Diamond | None | −13 dB (sinc) | Short-range, low clutter |
| LFM chirp | Tilted ridge | Strong (linear) | −13 dB (unweighted) | General-purpose search |
| Barker-13 | Thumbtack | Minimal | −22.3 dB | Precision range |
| Polyphase (P4) | Near-thumbtack | Low | −30+ dB | Low-sidelobe search |
| OFDM radar | Grid | None (orthogonal) | Depends on windowing | Cognitive/joint comms |
Frequently Asked Questions
What is the Ambiguity Function?
The AF |χ(τ,ν)|2 describes the matched filter output versus time delay (τ=range) and Doppler shift (ν=velocity). Woodward's theorem proves the total volume is constant: suppressing sidelobes in one region pushes energy elsewhere.
Why does LFM have range-Doppler coupling?
LFM sweeps frequency linearly. A Doppler shift is indistinguishable from a time delay at the matched filter input, producing range error ΔR = c·fd/(2·chirp rate). Nonlinear FM or OFDM waveforms break this coupling.
Can AESA radars change waveforms pulse to pulse?
Yes. Digital waveform generators switch waveforms per pulse, enabling cognitive radar: wide LFM for detection, Barker for range precision, CW for Doppler refinement, all within a single CPI.