Source Number Estimation using Hybrid Analog-Digital Array based on Spatial Covariance Matrix Reconstruction

The utilization of hybrid analog-digital arrays is prevalent in millimeter-wave band applications due to the advantages of lower complexity and cost compared to full-digital arrays. However, existing source number estimation algorithms cannot be directly applied to hybrid analog-digital arrays since the spatial covariance matrix (SCM) cannot be obtained. To address this issue, we propose a method that reconstructs the SCM through beam sweeping and subsequently applies classic source number estimation algorithms such as Akaike Information Criterion (AIC), Minimum Description Length (MDL) and Gerschgorin Disk Estimator (GDE). Our simulation results demonstrate that the proposed method performs comparably with using a full-digital array. Among the classic source number estimation algorithms, GDE exhibiting superior performance at both low and high SNRs.