Estimation of the parameters of complex exponentials under radar land clutter and single snapshot observation using Modified Forward Backward Linear Prediction

Estimation of the frequencies, amplitudes and phases of complex exponentials that model the targets present in radar data under additive white Gaussian noise and land clutter, is investigated stating Maximum Likelihood (ML) and employing spectrum estimation techniques. Use of Tufts & Kumaresanpsilas Modified Forward Backward Linear Prediction (M-FBLP) [D.W. Tufts and R. Kumaresan, 1982] in the defined problem, to the best of our knowledge, seems unpublished. The essential part of the work is that line spectra estimation techniques model the clutter process also as a complex exponential. To estimate the amplitudes and phases given the frequency estimates, two methods are employed and compared. Cramer-Rao bounds are also provided. It has been concluded that for the selected SNR range, M-FBLP shows better performance compared to others. In addition, it is stated that the optimum linear time invariant filter that can be placed before the M-FBLP deteriorates the operation.