Convolutional Neural Network Based Joint Estimation of Direction of Arrival and Polarisation Parameters

Abstract

In this paper, a convolutional neural network (CNN) based algorithm for joint estimation of direction of arrival (DOA) and polarisation parameters is introduced. Previous algorithms for the joint estimation of DOA and polarisation parameters frequently exhibit performance degradation under low signal-to-noise ratio conditions and with a limited number of snapshots, which imposes significant constraints on their practical applicability. To address these limitations, this paper proposes a CNN-based algorithm for the joint estimation of DOA and polarisation parameters. This method splits the joint estimation of DOA and polarisation parameters into two steps. In the first step, the upper triangular array of array output covariance matrices is used as data input to CNN to realise DOA estimation from signal sources. The second step establishes the polarisation and spatial domain spectral functions, followed by a spectral peak search to estimate the polarisation parameters of the signal. Simulation results demonstrate that the proposed algorithm achieves a significantly higher estimation accuracy compared to classical estimation methods under the conditions of low signal-to-noise ratios and a limited number of snapshots. The analysis of the simulation results shows that the proposed algorithm leverages a CNN, which fully exploits the spatial characteristics of the array output covariance matrix and consequently reduces the computational complexity.