The cell averaging constant false alarm rate detector has been assumed
to be optimal for detecting Swerling I targets embedded in exponentia
l clutter and noise of unknown power. This is because the detector use
s a minimum variance unbiased estimate (which is also a maximum likeli
hood estimate) of the unknown clutter-plus-noise power to set the thre
shold. In this paper, we prove, using a result concerning least favora
ble distributions in composite hypotheses testing, that the cell avera
ging detector is indeed optimal in that it is a uniformly most powerfu
l detector.