PE1-3.5 Adaptive Thresholding & CFAR Detection
You can explain why a fixed threshold fails in non-stationary clutter, describe the principles of adaptive CFAR detection and target models, and compare standard CFAR detectors.
A fixed threshold is only safe if the noise level never changes, which in radar it always does — clutter varies with terrain, weather and angle. CFAR solves this by estimating the local noise level from neighbouring cells and setting the threshold relative to it, keeping the false alarm rate constant as the name says. The detector variants (cell-averaging, greatest-of, ordered-statistic) differ in how they estimate that background, and each fails in a different way at a clutter edge or with multiple targets.
Work through these
Why fixed thresholds fail in varying clutter
Why a threshold set in advance stops working when the background changes. It is the practical problem the rest of this topic solves and it is worth understanding before the solution.
Principles of adaptive CFAR detection
Setting the threshold from the data itself so the false alarm rate stays put. The name says what is held constant, which is the whole design goal.
Target models
What you are looking for, modelled statistically. Whether the target fluctuates between looks changes the detector's performance substantially.
Cell-averaging CFAR
The standard method: estimate the background from neighbouring cells. Understanding why the cell under test is excluded is the key detail.
Review of other CFAR detectors and their failure modes
The variants, and where each of them fails. This is the useful half: the standard method breaks at clutter edges and with multiple targets, and each variant fixes one of those.
Sign in to keep your progress.
Free resources
We haven't checked most of these for screen reader use yet.
Links last checked 30 Aug 2026.
Stuck here?
Ask a mentor. A real person answers, and they can see exactly which topic you're on. Usually within a couple of working days.
Checking your session…
Topics shown in module order.