Abstract:

The estimation performance of traditional polarization-direction of arrival (DOA) parameter estimation methods will deteriorate seriously or even fail when some array elements fail.Meanwhile,error data from some failed array elements are introduced in the face of an increasingly complex electromagnetic environment.It is also a significant challenge to the robustness of the methods.For the above issues,to fully explore the impact of some array element failure and error data on the parameter estimation performance of the methods,two cases of partial elements complete failure and partial elements error probability are considered in the conformal polarization sensitive array (CPSA),and a spatially two-dimensional joint sparse polarization-DOA parameter estimation method based on variational sparse Bayesian learning (VSBL) is proposed.First,a two-dimensional sparse received array signal model based on the CPSA is established by using spatial sparse characteristics of the source.Second,the singular value decomposition method is used to reduce the dimension of the array output matrix,so as to reduce the computation load of the method.Subsequently,a robust DOA estimation is obtained by using the VSBL.Finally,the polarization parameter estimation of the source is obtained by the modulus constraint method.Simulation results validate that the proposed method has a relatively more robust parameter estimation performance and a higher estimation accuracy and a higher angle resolution in the case of array element failure.

Key words: signal processing, direction of arrival, sparse reconstruction, variational sparse Bayesian learning