Algorithms For Digital Micro Wave Receivers And Optimal System Identification PDF Download

The research in the Year-1 of this project has focused on two primary directions, as originally, proposed, (i) Advanced signal processing algorithms for digital microwave receivers with Electronic Warfare applications: For estimating the Angles-Of-Arrival or Radio Frequencies, a significant contribution has been made with a computationally efficient Minimum-Norm Method that does not require any Eigenanalysis but produces equally good estimates. A Maximum-Likelihood Estimator (MLE) that ensures unit circle frequencies has been proposed for obtaining the most accurate estimates. Furthermore, two new algorithms for improved AR/ARMA spectrum estimator from noisy observations have been considered. Several time-domain and frequency-domain algorithms for detecting the presence of targets are also being studied. (ii) A general and unified theoretical framework for optimal identification of rational transfer function coefficients from: (1) Input-Output data, (2) Impulse Response data and (3) Frequency Response data. Unlike existing algorithms which either modify or linearize the error criterion to estimate the unknown parameters simultaneously, the true error criteria have been decoupled into (i) a purely linear problem for estimating the optimal numerator and (ii) a nonlinear problem with reduced dimensionality for the optimal denominator. The decoupled estimators possess global optimality properties but have reduced computational complexity than existing methods. Angles-of-Arrival estimation, Frequency estimation, Digital receiver design, Improved AR and ARMA modeling, Electronic Warfare (EW) signal detection, Optimal system identification from input/output and frequency domain data.