FMCW LIDAR

# FMCW vs. Pulsed LIDAR Ranging Method

Pulsed LIDAR measures distance the straightforward way — send a short laser pulse, time exactly how long it takes to return, and convert that round-trip time directly into distance — while FMCW LIDAR instead transmits a continuous laser beam whose frequency sweeps steadily over time, and extracts both distance and velocity by comparing the frequency of the returning light against the frequency currently being transmitted. A pulsed system's distance measurement is as direct as it gets: time-of-flight divided by the speed of light gives distance, full stop, which is exactly why pulsed LIDAR (paired with the SPAD timing discussed in an earlier round) has been the mainstream automotive LIDAR approach. FMCW instead continuously chirps its transmitted frequency up and down; because light takes time to travel out and back, the returning signal's frequency at any instant reflects what was being transmitted slightly earlier — comparing the two frequencies directly yields distance, and because a moving target also Doppler-shifts the returning frequency, that same comparison simultaneously yields the target's velocity.

FMCW's direct velocity measurement is a real capability pulsed systems don't get for free: a pulsed LIDAR has to infer velocity indirectly, by comparing the distance measured on one scan against the distance measured on a later scan and dividing by the time between them, while FMCW measures velocity instantaneously within the same single measurement that produces the distance reading. Inferring velocity from successive distance measurements works, but it's inherently a derived, lower-resolution estimate limited by how far apart in time those two measurements were taken. FMCW's Doppler-based velocity comes directly from the physics of the measurement itself — no second measurement or time-differencing step required — which is particularly valuable for safety-critical automotive applications where knowing not just where an object is but how fast it's approaching matters as much as the distance itself.

FMCW's practical cost is coherent detection complexity: because the measurement depends on precisely comparing the returning light's frequency against the currently-transmitted frequency, FMCW systems need a highly stable, narrow-linewidth laser source and coherent receiver electronics — a meaningfully more demanding optical and electronic design than a pulsed system's comparatively simple transmit-pulse-and-time-the-return architecture. A pulsed system just needs to generate a sharp pulse and time its return accurately; an FMCW system needs its laser's frequency sweep to be precisely controlled and coherent enough that the frequency-difference measurement actually means something, which is a fundamentally harder laser-source and signal-processing problem to solve reliably and cheaply at automotive production volumes.

The MEMS scanning mirror discussed in the prior round determines where the beam points; FMCW versus pulsed is instead about what that beam fundamentally is and how its return is interpreted — spatial direction and ranging method are genuinely independent design choices, and a working LIDAR system has to get both right regardless of which specific combination it ultimately uses.