Parameter Estimation Using a Triad of Electrically Long Dipole Despite Data Model Mismatch

Most dipole-antenna signal processing algorithms have focused on "short dipole" whose physical length ($L$) is under ($1/10$) of a wavelength ($\lambda$). Recent works now handle "long dipoles" of an electrical length $L/\lambda \in [0.1, 1]$. This article advances an algorithmic paradigm that uses short-dipole algorithm on long-dipole data. By studying the mismatched Cramer-Rao bound, we find that applying short-dipole algorithm to a long-dipole data model may yield parameters' estimates that are not far from their true values especially when $L/\lambda$ is small. This phenomenon motivates a two-step algorithm where we apply short-dipole algorithm to get parameters' coarse estimates, which are then fine-tuned by an iterative algorithm. Our algorithm maximally retains the merits of the existing short-dipole algorithms while processing the signals received by a long dipole triad. Monte Carlo simulations verify the effectiveness of the proposed paradigm.