Radars listen to the echo of radio waves, like a microphone recording music. To make equipment cheaper, the 'recording' is done less frequently than theory demands: this is spectrally efficient (compressed) sensing. The cost of this economy is the emergence of false 'ghosts': the system sees targets that aren't there, just like off-key notes on old tape.
Engineers long chose between two 'melodies' of radio pulses. The linear one (Maxwellian wave with a straight frequency sweep) painted a clear picture but spawned phantom notes. The hyperbolic one removed ghosts, yet smeared close objects, like a choir where voices blend together. The new SC-EFM modulation changes frequency along a smooth arc. It's like singing a scale where each note stands on its own: even a cheap voice recorder can tell them apart. The idea was inspired by audio cassettes — on them, high tones suffer first, and the math of the new signal makes the radar insensitive to the 'wear and tear' of the recording.
Numerical experiments confirmed: the new 'melody' suppresses ghosts almost as well as the hyperbolic one, and distinguishes neighbors no worse than the linear one. Future radars, from parking sensors to orbital telescopes, will become sharper and cheaper.
as written by the authors arXiv:2607.14783
Sub-Nyquist sampling is an attractive way to reduce the hardware cost of wideband pulse-compression radar, but it introduces coherent alias-induced replicas in the matched-filter range profile, producing spurious peaks known as ghost targets. Existing frequency-modulated waveforms face a practical trade-off in this regime: linear frequency-modulated (LFM) pulses provide compact range responses but are highly susceptible to ghost-target detections, whereas hyperbolic frequency-modulated (HFM) pulses suppress ghosts at the cost of degraded target separability. To overcome this trade-off, we propose a sine-over-cosine Jacobi elliptic frequency-modulated waveform, referred to as SC-EFM, in which the elliptic modulus tunes the instantaneous-frequency (IF) curvature while preserving the pulse duration and bandwidth of conventional benchmarks. We characterize the sub-Nyquist folding structure of SC-EFM and derive closed-form expressions for the multi-target and ghost-target detection probabilities. Numerical results show that SC-EFM significantly suppresses ghost detections relative to LFM while matching its target separability, and substantially outperforms HFM in resolving close targets, providing a unified waveform solution for ghost-resilient sub-Nyquist multi-target ranging.
The first radar installation capable of detecting an aircraft appeared in 1935. Its pulses lasted microseconds, and its range did not exceed a couple of dozen kilometers.
In sci-fi sagas, ship sensors flawlessly distinguish every particle. Our radars with 'anti-ghost melody' are a step toward that same sharpness.