Study On Methods Of DOA Based On The Electromagnetic Vector Sensor Array

Array signal processing is a new field of signal processing, and it is a very important content of modern signal processing. Array signal processing has wide application in the field of national economy, military affairs such as communication, radar, sonar, navigation, geological exploration, mechanical and electronic measurement, biomedicine, radio astronomy and so on.Compared with one dimension signal processing of traditional single sensor, array signal processing has lots of advantages such as adaptive beamforming, strong noise suppression, super resolution power and so on. These advantages are the main power of the development of array signal processing.As we all known, for spacial electromagnetic signals, directions of propagation and polarized states are important characteristic parameters, they carry much important information. Directions of propagation describe the spacial positions of signal sources; polarized states describe the vector characters of electromagnetic waves' movement which are their properties of electromagnetic waves. Electromagnetic vector sensor arrays could not only gain the information of DOA, but also get the information of polarization. Full electromagnetic informations provide the physical foundation for promotion of the array performance; make it possible to increase the performance further. Compared with traditional scalar sensors, electromagnetic vector sensor array has a better performance: steady detective ability, stronger anti-jamming ability as well as higher space resolution.The element of electromagnetic vector sensor array can obtain the integrated electric and magnetic fields information or part of them, the perfect situation is that the element obtains both three orthometric electric informations and three orthometric magnetic informations, but between them there is some redundant information. On the condition of directions of propagation were known, as the electronic field is perpendicular to the direction of propagation, the three electronic field components are not independant with each other,it's the same as three magnetic field components.Besides, because of the multiplication cross relationship among the electronic field vector, the magnetic field vector and the Poynting vector, the electronic field vector and the magnetic field vector are not independant, we may choose parts of the dipoles ,at least two in order to obtain the integrated polarized information,to form the vector sensor.Usually we suppose the signals are generalized stationary, but in some actual circumstances, the signals are not stationary. Actually most artificial signals are non-stationary signals which change periodicly, so called cyclicstationary signals. Circular vector sensor array has many good characters compared with linear vector sensor array, for example, circular vector sensor array can estimate the azimuth angle and the polarization angle at the same time. The third chapter proposed Cyclic MUSIC method combined with 4th-order cyclic cumulant to estimate DOA, which based on UCA and cyclic- stationary signals. This method utilized UCA's manifold and the restraining-noise character of cyclic cumulant. Though the method needs large volumes of search, it can restrain colored noise well.In actual application, the estimation of coherent signals is one kind of important issues. This paper proposed polarization smoothing method for the DOA estimation of cyclostationary coherent signals based on the L electromagnetic vector sensor array, which solved the problem of coherent signals' DOA estimation. It also increases the number of coherent signal sources and does't effect the aperture. In the procedure of computing the parameters match each other automatically, this method utilized the signals' character of cyclic cumulant, so it has the ability of restraining interference signals and noise.