Publications Repository - Gdańsk University of Technology

To make decisions, cognitive radar must rely on predictions of its own performance. In the literature, these predictions are usually based on some form of Cram\'er-Rao lower bound. This approach is scientifically sound, but it also brings a possibility of the cognitive controller overestimating radar performance. It therefore makes sense to back theoretical predictions with careful experiments which will verify their applicability. Using a simple direction of arrival estimator as an example we discuss how an experimental performance curve can be obtained. We also propose an extended approach, which employs experimental data to estimate parameters of a performance curve which includes floor and threshold effects.

Authors

• dr hab. inż. Michał Meller link open in new tab