| Wireless localization technology is a method to determine the geolocation of mobile terminal within a certain range. In most of the applications, the first task is to estimate the DOAs of incoming signals. This estimate of the DOA can be used to localize the signal sources.The estimation of the direction of arrival angle is one of the most important research aspects in array signal processing. In the past thirty years, the scholars of the world had done a lot of research on DOA estimation of far field and achieved plentiful fruit.In one-dimensional (1-D) DOA estimation methods, the resolved directions of arrival are only located in a plane, but in application Systems, the sources are not in a plane, and then they would be expressed in two-dimensional angles.Two-dimensional (2-D) DOA estimation has played an important role in array signal processing in many areas such as Radar, sonar, and mobile communication systems.In this dissertation, by means of second order statistics and higher order statistics, combined with the knowledge of array signal processing and matrix theory, concerning key problems, example for the 2-D angles pairing, computation cost, resolution capacity, practicability, many methods of far-field narrow-band 2-D DOA estimation which is a typical multi-parameter estimation are thoroughly studied.The creative wokt of this dissertation as follows:1. Using the aperture extension property of higher-order cumulant, based on virtual array extending of the arbitrary array, a new method of DOA estimation is proposed. The coordinate and steer vector of virtual sensors are computed from the coordinate of actual sensors, using disparate virtual sensor to construct the higher-order covariance matrix, then using MUSIC algorithm to estimate DOA.2. Introduce Matrix Enhancement and Matrix Pencil (MEMP) to estimate 2-D direction of arrival angles. To overcome the shortage of MEMP method, based on L-shape array, utilizing the output of linear array to calculate the cross covariance matrix or cross forth-order cumulant matrix, using MEMP method to construct enhanced matrix. The 2-D MEMP DOA estimation method, extended MEMP DOA estimation method and 2D-MEMP-RMRLA-D0A estimation method are proposed respectively. The method is effective in rapid estimation and high-resolution of 2-D DOA angles, improves the performance of the 2-D DOA estimation.3. Proposed a space-time algorithm to estimate the azimuth and elevation angles of multiple independent sources based on arbitrary array. The algorithm transfers the observed data from space domain to 2-D space-time domain by exploiting the cross-correlation of the array outputs, constructs space-time cross-correlation matrixes. The algorithm only needs three guiding sensors, using 2-D ESPRIT method, by M sensors can estimate carrier frequency, azimuth and elevation angles of M sources which have different carrier frequency.The dissertation consists of eight chapters.In chapter one, analyzed and summarized the main methods of two-dimensional(2-D) direction-of-arrival(DOA), current research state and the significance of 2-D DOA estimation, the key problems in 2-D DOA estimation. Finally the research content of this dissertation is proposed.In chapter two, some fundamental knowledge which would be used later is introduced, such as matrix algebra, the fundamental knowledge of higher-order statistics, models of common planar arrays.In chapter three, researching the aperture extension property of higher-order cumulant, a new method of DOA estimation is proposed based on virtual array extending of the arbitrary array. The coordinate and steer vector of virtual sensors are computed from the coordinate , using virtual array to construct the higher-order covariance matrix, then using MUSIC algorithm to estimate DOA. The method can estimated DOAs based on the arbitrary array. Compared with MUSIC-like method, the method not only has the precision but also has low computation cost.In chapter four, based on uniform L-shape array, a novel method for estimating 2-D DOA is presented .The method uses the array geometries to calculate three cross correlation matrixes and uses those matrixes to construct a special large matrix, and then obtain the required the signal subspace in the 2-D ESPRIT method via the eigendecomposition of the constructed matrix. Without spectral peak seaching and complex computation, this algorithm works well for pairing among parameters. It is shown through the simulations that this method has high precision, low computation cost and some practicability.In chapter five, in the coherent signals case, based on the uniform rectangular array, proposed 2-D DOA estimation methods of coherent signals based on spatial smoothing and temporal smoothing. using 2-D spatial smoothing technique and EVESPA method, 2D-SS-EVESPA-DOA method is proposed for coherent signal. Using temporal smoothing technique and EVESPA method, 2D-SS-EVESPA-DOA method is proposed for coherent signal. Those methods utilize the 2D Unitary ESPRIT method which pairs the 2-D angles automatically. The method could group the 2D DOA angles according the different source, without spectral peak seaching and pair the 2-D angles automatically.In chapter six, introduce Matrix Enhancement and Matrix Pencil (MEMP) to estimate 2-D direction of arrival angles. Based on uniform L-shape array, by means of second order statistics, 2-D MEMP DOA estimation method is proposed, the method computes the covariance matrix of output of the two linear arrays and construct the enhanced matrix by MEMP method, and then estimate the 2-D DOA by the ESPRIT method. Using a new algorithm makes the 2-D angles pairing automatically. To avoid the aperture decrease, an extended MEMP DOA estimation method using the aperture extension property of fourth-order cumulant is proposed, the method increases the valid aperture of the arrays, without spectral peak searching. Based on Minimum- Redundancy Linear Arrays, 2D-MEMP-RMRLA-DOA estimation method is proposed, the method compute the covariance matrix of output of the two RMRLAs, construct the enhanced matrix and estimated 2-D DOA angles by MEMP method. The 2D-MEMP method is effective in rapid estimation and high-resolution of 2-D DOA angles, improves the performance of the 2-D DOA estimation.In chapter seven, Proposed a space-time algorithm to estimate the azimuth and elevation angles of multiple independent sources based on arbitrary array. The algorithm transfers the observed data from space domain to 2-D space-time domain by exploiting the cross-correlation of the array outputs, constructs space-time cross-correlation matrixes. The algorithm only needs three guiding sensors, and using 2-D ESPRIT can estimate carrier frequency, azimuth and elevation angles of multiple independent sources which have different carrier frequency.In chapter eight, a brief summary of the dissertation is given, and the future work is also expected.
|
PDF Full Text Request