Joint 2D-DOA and polarization estimation for Electromagnetic vector sensors array with compressive measurements

Electromagnetic vector sensors (EMVS) arrays have been employed extensively in the field of array signal processing for their advantage in terms of polarization diversity. However, with the introduction of EMVS, the cost of the hardware equipment and the complexity of the corresponding parameter estimation algorithms increase considerably, as the number of received signal channels and the dimension of the received signal are much larger than the traditional scalar array. In order to effectively reduce hardware cost and algorithm complexity, we propose a scheme that combines an electromagnetic vector sensor array with a compression network. We construct the corresponding signal model and based on this we derive a Compressed Reduced Dimensional MUltiple SIgnal Classification (Compressed To avoid the multi-dimensional search, Reduced Dimension MUSIC) algorithm which can effectively reduce the computational complexity. While selecting the coefficient matrix of the compressed network, random selection can cause information loss, which leads to the performance degradation of the estimation algorithm. To address this problem, we propose an optimization method for coefficient matrix selection based on the maximum signal-to-noise ratio (SNR) criterion. Numerical simulations are conducted in different scenarios to verify the effectiveness of the parameter estimation algorithm and the optimization algorithm.

Automated summary

This article introduces a scheme that combines electromagnetic vector sensor arrays with compression networks to reduce hardware cost and algorithm complexity. A compressed reduced dimensional multiple signal classification algorithm is derived based on the signal model to effectively reduce computational complexity. An optimization method based on signal-to-noise ratio criterion is proposed to address the information loss problem in coefficient matrix selection.