Abstract
Crossed-loop/monopole antenna has been widely used in high frequency ground-wave radar for ocean surface remote sensing owing to its compact size. The multiple signal classification (MUSIC) algorithm is commonly adopted to achieve a favorable angular resolution for this broad beam antenna. However, the direction of arrival (DOA) estimation is always troubled by the frequently occurring distortion of antenna pattern. Previous studies are mainly focused on the calibration by using the actual pattern or other assistant algorithms. In this paper, theoretical analysis is made to illustrate the reason for pattern distortion, which draws a conclusion that the distortion is essentially due to the different influence of environmental obstacles upon the electric field part and magnetic field part. This also implies that there is nearly no distortion between the two loops, which is validated by three different experiments. Based on this conclusion, the DOA estimation performance of MUSIC algorithm is studied by creating a numerical relationship associating the estimation error with distortion level. Furthermore, a new method using ideal pattern is proposed to improve the estimation performance by removing the amplitude information of monopole and relying on the crossed-loop processing. Both simulation and experimental results are given to prove its validity.
Funding Information
  • Ocean Public Welfare Scientific Research Project (201205032-3)
  • National Special Program for Key Scientific Instrument and Equipment Development of China (2013YQ160793)
  • Fundamental Research Funds for the Central Universities (2042016kf0019)

This publication has 24 references indexed in Scilit: