Research On Direction Of Arrival Estimation Methods In Multipath Environment

Direction of arrival(DOA) of spatial signal is an important concept in array signal processing, which including azimuth angle and elevation angle. A precise estimation of signal’s DOA is very important in a few fields such as radar, sonar and communication system, but some high resolution algorithms based on subspace will fail when signals are coherent. As signal’s multipath propagation is very common in real environment, one hot issue of array signal processing is how to estimate DOAs of the coherent signals due to multipath propagation exactly. Main work of this thesis is to research the DOA estimation methods in multipath environment, which can be summarized as follows:Firstly, time-domain model and frequency-domain model of the signals received by the array in the presence of mutlpath signals is derived, the correlation of two multipath signals is analyzed, and forward/backward spatial smoothing algorithm is introduced.Secondly, DOA estimation methods based on ordinary array in multipath environment is researched. The methods to realize the DOA estimation of coherent signals using the centrosymmetric structure of array are researched in the first place, and for circular array, application condition of the decorrelation algorithm based on centrosymmetric structure of array is analyzed. Then, spatial differencing method using oblique projection operator is researched, which can estimate the DOAs of uncorrelated signals and coherent signals separately, in this method, DOAs of coherent signals can be estimated in group using the virtual array manifold constructed according to the differentiated matrix. At last, in gain and phase error calibration, when the signal emitted by auxiliary source has multipath propagation, a method to estimate multipath interfering signal’s DOA in the presence of sensor gain and phase errors is studyed.Thirdly, DOA estimation methods based on four dimensional array in multipath environment is researched. Working principle of four dimensional array and DOA estimation method based on four dimensional array are introduced first. And then, DOA estimation method based on signal subspace fitting is studyed, which has the ability to handle coherent signals, modified Gauss-Newton method is used to solving the nonlinear multidimensional optimization problem. Finally, according to the array structure and time sequence of the four dimensional array with unidirectional phase center motion, a spacetime signal model suitable for single frequency signal is provided, but for narrowband signal, this model requires the signal is highly correlated in the whole sampling period of the four dimensional array, time smoothing of array’s sampling data in different time period is used to decorrelation, and then coherent signals’ DOA can be estimated using MUSIC algorithm.