Toward efficient super-resolution in SatCom: DoA estimation with hybrid antenna arrays

High path loss is a well-known challenge in Satellite Communications (SatCom). The emergence of Multiple-Input Multiple-Output (MIMO) systems offers a promising solution by enabling the synthesis of high-gain, highly directive beams. However, such beams require extremely accurate Direction of Arrival (DoA) estimation, as even small pointing errors can significantly degrade performance. This need is amplified in modern SatCom scenarios, particularly with the renewed interest in Low Earth Orbit (LEO) constellations, where receivers must often resolve multiple closely spaced satellites. Achieving such resolution demands the use of super-resolution techniques. Although DoA estimation is a well-established field, most super-resolution algorithms—such as MUSIC or ESPRIT—assume fully digital architectures and cannot be directly applied to Hybrid Analog-Digital (HAD) arrays, which are essential for reducing hardware complexity in large-scale antenna systems. To address this limitation, we propose an algorithm that estimates the second-order statistics of an equivalent fully digital array by exploiting the specific Toeplitz structure of the covariance matrix. This enables the use of classical DoA methods in HAD architectures, making super-resolution feasible in practical SatCom systems.