WiFi sensing - RF front-end requirements vs communication?

WiFi sensing reuses communication hardware but needs better phase stability. Did your existing front-end meet sensing requirements, or did you have to upgrade the RF chain?

WiFi sensing (CSI / Doppler / presence) reuses a communications RF chain, but the front-end requirements diverge once you care about stable phase and repeatable amplitude across packets.

What usually matters beyond a stock WiFi NIC:
1. Phase stability: sensing wants coherent CSI across time; crystal / PLL wander, AGC steps, and chain-to-chain skew show up as fake motion. Comms can ride AGC; sensing often needs frozen gain or logged AGC state.
2. Bandwidth vs delay resolution: wider channel bonding helps range resolution, but only if the analog front-end is flat and the calibration of IQ / cable delay holds across the band.
3. Antenna / multipath: sensing exploits multipath that communications equalizes away - fixed geometry and a calibration empty-room capture help more than a "better" PA.
4. TX leakage and self-interference: full-duplex-ish or high-duty beaconing for sensing stresses isolation and RX linearity differently than sparse traffic.
5. Sync: multi-AP or multi-antenna sensing needs a shared time/phase reference; without it you are doing non-coherent RSS heuristics, not RF sensing.

If you share whether you need presence, respiration-scale Doppler, or imaging-ish range profiles, and which WiFi generation / channel width, people can say how far a stock AP can go vs when you need a SDR / calibrated array.

— rfforum 编辑部 / Editorial Team