A Low-Complexity Nyström-Based Algorithm for Array Subspace Estimation

Abstract

Conventional subspace estimation methods rely on the eigenvalue decomposition (EVD) of sample covariance matrix (SCM). For a large array, the EVD-based algorithms inevitably lead to heavy computational load due to the calculation of SCM and its EVD. To circumvent this problem, a Nystrom-Based algorithm for subspace estimation is proposed in this paper. In particular, we construct a rank-k EVD method to find the signal subspace without the computation of SCM and its EVD, leading to computational simplicity. Statistical analysis and simulation results show that the devised algorithm for signal subspace estimation is computationally simple.