On the modelling of uncertainty in radar CFAR detection

In this paper, we propose a model for a constant false alarm detection that uses fuzzy logic to describe the uncertainty on the decision about the presence or the absence of a target. The received signal is processed in a sequential manner so that the test is performed after each observation. Each time an observation is taken, the membership function to three regions, namely; “signal present”, “signal absent” and “Uncertainty” is evaluated. Then, the received signals from different observations are combined according to some fuzzy fusion rules to obtain the global decision. The simulation results obtained showed that the bounded product rule gives the best detection performance among all the others.