The CLEAN algorithm is an iterative deconvolution technique originally developed in radio astronomy and adapted for radar and WiFi sensing to resolve closely spaced targets in range-Doppler or delay-Doppler profiles. It works by repeatedly identifying the dominant peak in a signal response, subtracting a scaled version of the point spread function centered at that peak, and accumulating the detected components, thereby suppressing sidelobes and improving target separability beyond the native resolution of the aperture or waveform. In WiFi and ISAC sensing contexts, CLEAN is significant because it enables accurate multipath component extraction and target localization even with limited bandwidth or antenna configurations, with key variants including the classical Högbom CLEAN, its application to synthetic aperture radar (SAR) imaging for through-the-wall scenarios, and extensions tailored to OFDM channel estimates for separating overlapping reflections from multiple targets or scatterers.
Source Papers
- Enabling ISAC on Low-Cost Devices via Spatial-Channel Estimation With a Single-RF Chain ↗ — Enabling ISAC on Low-Cost Devices via Spatial-Channel Estima
- OPERAnet, a multimodal activity recognition dataset acquired from radio frequency and vision-based sensors ↗ — OPERAnet, a multimodal activity recognition dataset acquired
- Passive WiFi Radar for Human Sensing Using a Stand-Alone Access Point ↗ — Passive WiFi Radar for Human Sensing Using a Stand-Alone Acc