Design and Simulation MUSIC Algorithm for DOA Estimation in Comparison to Bartlett and Capon Algorithms for Smart Antenna
Mohamed A. Alkelsh, Mohammed M Alshibany, Walid K. Hasan · Tobruk University Journal of Engineering Sciences · 2022
This paper describes three schemes to implement Direction of Arrival Estimation (DOA) that is used and implemented in real Antenna system. The variation between those algorithms is important in means of accuracy and complexity of implementation such mathematical models in actual system of Antenna. The use of mathematical analysis is very helpful to understand the process of DOA estimation and each part of the algorithms in the study, so that can be used further to implement algorithms and work on MATLAB to investigate and evaluate the performance of each one in relative with the others and how they act when they put in same conditions with different number of samples and using different factors. The used algorithms for the study are the conventional algorithms Bartlett & Capon, in contrast with more advanced algorithm MUSIC Multiple-Signal-Classifications, which is introduced as the main focus in the study. Simulation of DOA is made using MATLAB and the mathematical implementation of the three algorithms. The results of comparing the algorithms shows that varying number of samples taken has a crucial role on determining Direction of Arrival or the incident angle of the input stray-signal at the arrays of the smart antenna which senses the direction based on the algorithm operated in the smart antenna system, the use of the three algorithms in different number of samples had shown a performance change of accuracy of estimation in different way for an algorithm in comparison to another algorithm in many circumstances, the more prevailing fact is that the more number of samples the more accuracy of estimation, but in case of less samples the Bartlett algorithm showed better performance in compare to other algorithms, as for Capon algorithm it suffers accuracy at low number of samples and need high number of samples to act more accurately, as for MUSIC algorithm it shows the best overall performance as adapting fast to the angle of estimation with high accuracy in response to an efficient number of samples.