Effectiveness and Characteristics of Virtual Antennas in the Multiple Signal Classification Algorithm

• DOI: 10.3390/electronics14010073
• Publication Date: 2024-12-27T14:23:36Z
• Authors: Riku Takemoto; Jaesang Cha; Incheol Jeong; Chang-Jun Ahn
• Keywords: 0203 mechanical engineering; 02 engineering and technology

This paper proposes a novel direction of arrival (DoA) estimation method based on the Multiple Signal Classification (MUSIC) algorithm using virtual antennas. The MUSIC method is a widely used DoA estimation technique known for its high accuracy and high resolution. However, it has a fundamental limitation: when the necessary condition of having more antenna elements at the base station than users is not met, DoA estimation becomes infeasible. To address this limitation, we introduce the concept of virtual antennas, allowing DoA estimation, even under restricted conditions. Virtual antennas are not physically present but can be virtually arranged, enabling the generation of virtual-received signals similar to those of real array antennas. Through simulation experiments, we demonstrate that our approach enables DoA estimation with the MUSIC algorithm even when its necessary conditions are not satisfied. Additionally, this paper explores the characteristics of virtual antennas in detail. We conduct simulation experiments to examine the differences in estimation accuracy between real and virtual antennas, as well as the impact of virtual antenna arrangement and count on estimation accuracy. The results show that, although virtual antennas provide lower estimation accuracy compared to real antennas, their flexible arrangement allows for improved resolution when signal sources are closely spaced by increasing the spacing between virtual antennas. Furthermore, under Additive White Gaussian Noise (AWGN) conditions, increasing the number of virtual antennas enhances estimation accuracy.