Study On Multidimensional Parameter Estimation Based On Non-ideal Electromagnetic Vector Sensor Array

The electromagnetic vector sensor array signal processing which has been developed rapidly in recent years, is a very important content of array signal processing and used widely in application such as radar, sonar and communication etc. Multidimensional parameters estimation of signals using array of electromagnetic vector sensors has become a hotspot.Electromagnetic vector sensor(EVS) can acquire the polarization information besides the spatial information of the Electromagnetic signals,which consists of a group of electromagnetic sensitive elements.EVS outperforms the traditional scalar sensors due to the additional polarization information.EVS possesses strong anti-jamming ability, robust detecting performance, high resolving power and polarization multi-address capability. All of these virtues make the the electromagnetic vector sensor array widely used in military and civil area. The electromagnetic vector sensor array signal processing mainly consists of two parts: signals filtering and multi-parameters estimation of electromagnetic vector sensor array. The aim of the signals filtering of electromagnetic vector sensor array is to suppress the interference signals and to enhance the expected signals by using the difference in spatial-field and polarization-field of the interference signals and the expected signals, and the precondition of the signals filtering of electromagnetic vector sensor array is that the spatial and polarization information of the signal must be known. The aim of the multi-parameters estimation of electromagnetic vector sensor array is to estimate the direction of arrival and polarization parameters and so on based on the electromagnetic vector sensor array withoutany prior knowledge if the spatial and polarization information of the signal. This paper mainly discusses the problem of multi-parameters estimation of electromagnetic vector sensor array.At present the mass of method used in electromagnetic vector sensor array signal processing domain is based on linear array. But the UCA attracts people's attention because of its excellent properties. All of the existing electromagnetic vector sensor array signal processing algorithms assumes that each antenna of the electromagnetic vector sensor orients the referenced Cartesian coordinate strictly, but in real application, the above approximation is not always accurate. The unknown orientation errors are inevitably exist, which depresses the estimation performance of the existing methods greatly.The electromagnetic vector sensor array has rich redundant information. Using this characteristic and with the concrete characteristic of the signal, we research the estimations of Multidimensional Parameter based on non-ideal electromagnetic Vector Sensor Array.Firstly, we discuss the DOA and polarization estimation based on the UCA, and the following methods are proposed: (1) Two-dimension DOA and polarization estimation of DOA matrix method. By sufficiently utilizing the geometry property of the UCA and the polarization invariance between the UCA, the proposed method gives the closed-form estimation through the eigendecomposition of the constructed DOA-matrix; (2) Forth-order cumulant based DOA, frequency and polarization estimation method. This method is suitable in arbitrary Gauss noise without any frequency information, and gives robust DOA estimation under the array perturbations environment. Simulation results show that these algorithms have good performances.Secondly, we discuss the estimation and calibration problem of the electromagnetic Vector Sensor Array orientation errors, using a calibration source which DOA and polarization parameters are known , following methods are proposed: single auxiliary source calibration method of electromagnetic vectorsensor and electromagnetic vector sensor array, moreover discuss the choice of parameters for calibration source, and give the corresponding CRB for each parameter. Simulation results show that these algorithms have good performances.