Insect Symmetry-Driven Orientation Estimation for Entomological Radar Using Multifrequency Scattering Matrices

Abstract

Entomological radar utilizes full-polarization data to estimate insect orientation, which is essential for understanding the orientation mechanisms of migrating insects and predicting their trajectories. Traditional orientation estimation methods rely on the empirical assumption that maximum echo intensity occurs when the polarization direction aligns with the insect’s body axis. Orientation is then extracted by identifying the polarization direction corresponding to the maximum echo intensity, based on the polarization pattern or the scattering matrix (SM) measured by single-frequency radars. However, the accuracy of the estimated orientation is affected by noise and polarization errors. To further improve orientation accuracy, based on a new generation of multifrequency and full-polarization entomological radar, this article proposes a novel method. The approach integrates multifrequency SMs of an insect under the assumption of insect body symmetry. First, a parametric SM model, characterized by four independent parameters, including insect orientation, was developed based on the polarization theory that when the symmetry axis of a symmetric target aligns with the horizontal or vertical polarization direction, the cross-polarization elements in the SM are zero. Using this principle, a cost function was constructed by summing the cross-polarization powers across multifrequency SMs. By minimizing the cost function, the analytical formula for insect orientation estimation was derived. Simulations using data from 159 insects measured in an anechoic chamber, along with field measurements, demonstrated that the proposed method provides superior accuracy and robustness against noise and polarization errors compared to traditional single-frequency approaches.