Sensitivity study of GLRT detection in compound-Gaussian clutter

We present a sensitivity analysis of a GLRT-based (generalized likelihood ratio test) processor designed to detect coherent pulse trains in Weibull-distributed clutter-dominated environment. More precisely, this work covers both the cases of fluctuating targets plus clutter with known statistics and non-fluctuating targets plus clutter with statistics not perfectly known. The sensitivity analysis shows that the receiver is relatively insensitive to possible mismatching between the design and the actual model of both the useful target and the clutter, thus demonstrating the suitability of the proposed strategy as a means for circumventing the a priori environmental uncertainty.