2-D high precision DOA estimation based on SVM

Nowadays, high-resolution direction of arrival (DOA) estimation techniques have been widely used in many fields such as sonar, communication, radio astronomy and biomedicine etc. Especially in the future wireless communication, base station (BS) is equipped with massive antennas to perform two-dimensional (2-D) DOA estimation. Unfortunately, conventional algorithms cannot achieve high precision due to the limitation of the array aperture and usually suffer from high computational complexity. Thus, we innovatively propose a SVM-based one-bit DOA estimation algorithm for 2-D DOA estimation. The 2-D DOA estimation problem is modeled as a classification problem due to one-bit quantification. Then, an iterative refinement procedure based on grid refinement is presented to achieve more accurate DOA estimation. Finally, simulations illustrate that the proposed method outperforms the existing methods in 2-D DOA estimation, effectively avoids the process of parameter pairing and requires less antenna elements.