Limitations of crosscorrelation radiometer performance

Conventional analyses of the crosscorrelation radiometer are based on an ideal assumption of a perfectly correlated signal, and Gaussian noise that is uncorrelated between the sensors. In the present work, the author predicts crosscorrelation radiometer performance under less ideal assumptions: non-Gaussian noise, nonzero noise correlation, and imperfect signal correlation. Using a noise model that accounts for uncertainty in the noise and signal parameters, he shows that the crosscorrelation radiometer is sensitive to uncertainty in the noise correlation coefficient. Moreover, even if the noise correlation coefficient is known exactly, detection performance degrade when there is uncertainty in the noise power