RF harvesting from ambient cellular signals - microwatts enough?

Ambient cellular energy sounds free but the power levels are microwatts at best. Has anyone powered a real sensor this way, or is it still a demo that dies the moment you add a radio duty cycle?

Ambient cellular harvesting can blink an LED or trickle a capacitor; powering a sensor *plus* a radio duty cycle is where most demos die.

Reality check:
1. Available power in non-absurd locations is often microwatts to low tens of microwatts after the rectifier - fine for ultra-low-duty sensing, not for chatting on LTE/Wi-Fi every few seconds.
2. Budgets that work: harvest into a storage element, wake rarely, send a short burst, sleep hard. The radio energy per uplink usually dominates; sensing and MCU sleep current are the other killers.
3. Location matters more than antenna cleverness: near a macro / indoor DAS you might scrape more; deep indoor / rural you get almost nothing useful.

If you have a real deployment (not a bench demo), posting average harvested power, wake interval, and bytes per uplink is more useful than another rectifier schematic. Share band and whether the node has a battery assist - that usually separates "harvester" from "battery with a marketing antenna".

— rfforum 编辑部 / Editorial Team