Research Of Polarization-DOA Estimation Algorithm Based On Quaternion

Extraction of the location of emitter from signals collected using an array ofsensors is known as the estimation of direction of arrival. Estimation of DOA couldapplied to wireless communication, radar, radio astronomy, sonar,navigation, trackingof various objects, and other engineering applications. Electromagnetic vector-sensorarray has better system performance, higher anti-angular fuzzy ability, higherrobustness of model error then scalar sensor array, also has polarized multiple accessability, which made it get more attention in wireless communication and informationwars. As a hypercomplex constructure, quaternions can obtain a more accurate modelof VSA. For this reason, based on VSA and quaternions, this paper aims to studypolarizatoin-DOA estimation algorithm based on quaternions, summarized as follow.Firstly, this paper summarizes related matrices and algebras, details theory ofquaternions and matrix of quaternions, and obtains theoretical basis of the excellentperformance of using quaternions through analyzes the orthogonality of vectors inquaternions field. Model in complex field just piles up of inner contents ofelectromagnetic vector-sensor and could not characterizes about the vector relationof them properly through analyzes the model of electromagnetic vector-sensors incomplex field and the corresponding V-MUSIC algorithm.Secondly, this paper models the electromagnetic vector-sensors array inquaternion field. Then introduces a polarization-DOA estimation algorithm based onquaternions(DR-MUSIC), through analyzes interior of quaternion arithmetic, andextracts parts which include significance information, this paper first proposes thehalf-quaternions algorithm(HQ-MUSIC). Then analyzes complexity of above threealgorithm through covariance matrix and and peak search in time consuming andmemory usage and the robustness of them when coherent noise exists. The resultproves that algorithms based on quaternions are less complexity than V-MUSIC. Lastwe test and verify the performance of the algorithms through simulation.Finally, this paper considers the scene of coherent signals. First analyzes spatialsmoothing algorithm based on quaternions(QSSA). Modeling in quaternion fieldaffects polarization steering vectors, this paper solves the problem of that model inquaternions field would halve the number of polarized steering vectors, then firstproposes the polarization smoothing algorithms based on quaternions applicable forthe two-component electromagnetic vector-sensor by constructing a new set ofpolarization steering vector, proposes the polarization smoothing algorithms based onquaternions applicable for the full six-dimensional electromagnetic vector-sensor bydividing array into polarized sub-arrays. Last we test and verify the performance of the algorithms through simulation.