Joint Estimation of Location and Polarization in Direct Position Determination Using a Moving Single Station

This paper studies the problem of direct position determination (DPD) of a moving single station. It is well known that the DPD methods do not require the estimation of intermediate positioning parameters, but directly process the original sampled signals to determine the position of the source. Furthermore, the DPD methods can locate multiple sources simultaneously, avoiding the data correlation problem in traditional twostep localization methods. Motivated by this, a DPD method based on joint estimation of location and polarization is proposed, and the cost function is constructed by using dimensionality reduction multi-signal classification (DRMUSIC) method. The first step involves obtaining the estimated covariance matrix of the received data through maximum likelihood estimation. Next, perform eigenvalue decomposition of the covariance matrix to obtain the noise subspace. Based on the fact that the polarization vector is non-zero, dimensionality reduction is achieved using the Lagrange multiplier method. The position parameters of the source are obtained by searching for the two-dimensional spectral peaks. Additionally, estimates are made for the polarization auxiliary angle and the polarization phase difference. The proposed algorithm achieves high positioning accuracy and resolution for adjacent sources. The simulation results demonstrate the effectiveness of the proposed algorithm.