7.b. CFAR Flashcards
Flase Alarm Rate
Number of false alarms occuring within a time interval.
What does a CFAR detector do?
Estimates the properties of interference form radar measurements, and adjust the detection threshold accordingly to maintain a constant falase alarm rate.
Explain how CA-CFAR works
Cell Averaging CFAR
Each bearing is split into range bins. For each CUT (Cell Under Test), the average power of a number of leading and lagging cells on either side of the CUT are averaged and compared to the CUT to determine a unique Detection Threshold. There are G (Guard) cellls on either side of the CUT so that a large target which takes up many cells does not overlap with the Leading and Lagging windows.
What is a limitation of CA-CFAR
Only works in a homogenous environment.
Homogenous vs. Heterogenous Environment
Geterogenous:
Target returns are present in leading/lagging windows and target is also present in the CUT
Interference sources are not identically distributed throughout the reference window
Target Masking
If there are two targets close together, CA CFAR may detect one of the targets in its reference windows and increase the threshold detection. This may lead to the threshold increasing past the target power so one of the targets is obscured in the reference window.
OR this may describe when there are no guard cells around the CUT, causing a large target to overlap with the reference window, raising the threshold so that the target is masked.
Clutter Boundary
If there is a sudden spike in clutter, the CFAR threshold does not increase quick enough, leading to a false alarms.
Greatest-Of CA-CFAR
Works by computing the average interference power in the lagging and leading windows separately, and then selecting the larger of the two means as the CFAR statistic.
Reduces clutter boundary false alarms.
Worse for interfering targets
Pros and Cons of Greates Of CFAR
Reduces clutter boundary false alarms
Exhibits degrated performance with interfering targets.
Smallest-Of CA-CFAR
Estimates interference power in lagging and leading referencewindows and selects the smaller of the two estimates.
Suprresses interfering targets that may reside in either the leading or lagging window.
Pros and Cons of Smallest-Of CA-CFAR
Increased clutter boundary false alarms.
Better performance with interfering targets.
Censored CFAR
Rank orders all measured samples in the reference window, discards the largest samples (which may contain returns from interfering targets.
Trimmed Mean CFAR
Discards the largest and smallest samples in the reference window.
Ordered Statistics CFAR
Ordered Statistics CFAR
Orders the N samples in the reference window and selects the k-th example as the CFAR statistic.
Rejects N-k interfering targets and suppresses clutter edge false alarms if k>N/2.