Optimal detection and estimation in FMCW radar

The paper is mainly concerned with detection of an accelerating target echo by FMCW radar. The processed signal is modeled as a polynomial phase signal, contaminated either by white or colored additive Gaussian noise. Obtained tests, optimal in the Neyman-Pearson sense, are linked with the maximum likelihood estimation procedures dealt with in earlier papers. An example shows the importance of properly suiting the test to the noise statistics.