Procedures for Super-resolution Extraction of nearby Targets by Radar with Multiple Input and Multiple Output

Numerous research have been conducted during the past 20 years on the challenge of detecting the directions of many spectrum signals from data received by a large number of sensors. One of the earliest methods was beam-forming, whose resolution is constrained by the array construction. At weak signal-to-noise ratios, this methodology was unable to identify collinear arriving angles. Additional method for parameter extraction of nearby targets is the maximum likelihood (ML) solution. Even though multidimensional optimization problem calls for a lot of processing power, global convergence is not often guaranteed. These are the main issues with ML techniques. Recently, methods for super-resolution DOA estimation based on subset methods, such as classification of different signals by MUSIC, Rotational In-variance Techniques for Signal Parameter Estimation by ES-PRIT, and Fitting a weighted subset known as WSF, have received a lot of attention. In this paper, we discussed all of the above algorithms for arrival angle estimation of two or more nearly spaced targets with less number of snapshots in Radar-based systems. Existing direction of entrance (DOE) estimation methods in radar systems will perform well in the cases of sufficient snapshots and a better signal-to-noise ratio (SNR). Here, by considering closely spaced targets or strongly correlated sources, all algorithms are tested in Matlab. Through computer simulation experiments, the performances of the MUSIC, ESPRIT, and WSF algorithms were compared. From this experiment, come to know that MUSIC provides angular resolution with greater accuracy and stability as compare to other methods.