Constant False Alarm Rate

Constant False Alarm Rate (CFAR) is a target detection scheme that adaptively sets a threshold to maintain a constant probability of false alarm regardless of the local noise level. For each cell, the threshold is proportional to the average of a ring neighborhood around that cell, where the deadzone immediately around the given cell contains “guard cells”.

$${\hat{P}}_{noise}[k]=\text{avg}(\mathcal{R}(k)),\text{threshold}[k]=\alpha \cdot {\hat{P}}_{noise}[k]$$

• where $\mathcal{R}(k)$ is the ring of reference cells surrounding bin $k$, excluding the guard cells immediately adjacent to the $k$th cell.
• $\alpha$ is chosen to achieve the desired false alarm rate.
• CFAR assumes the noise floor is locally flat, so it performs poorly in spectrally shaped or wideband environments where reference cells may contain signal energy. The Automatic Noise Floor Spectrum Estimation in the Presence of Signals was developed to address this limitation.