Akademik Veri Yönetim Sistemi
Tezin Türü: Doktora
Tezin Yürütüldüğü Kurum: Orta Doğu Teknik Üniversitesi, Mühendislik Fakültesi, Elektrik ve Elektronik Mühendisliği Bölümü, Türkiye
Tezin Onay Tarihi: 2016
Öğrenci: AHMET MUSAB ELBİR
Danışman: TEMEL ENGİN TUNCER
Özet:In direction finding (DF) applications, there are several factors affecting the estimation accuracy of the direction-of-arrivals (DOA) of unknown source locations. The major distortions in the estimation process are due to the array imperfections, model mismatches and multipath. The array imperfections usually exist in practical applications due to the nonidealities in the antenna array such as mutual coupling (MC) and gain/phase uncertainties. The model mismatches usually occur when the model of the received signal differs from the signal model used in the processing stage of the DF system. Another distortion is due to multipath signals. In the multipath scenario, the antenna array receives the transmitted signal from more than one path with different directions and the array covariance matrix is rank-deficient. In this thesis, three new methods are proposed for the problems in DF applications in the presence of array imperfections, model mismatches and multipath. In the first problem, calibration of antenna arrays mounted on aeronautical vehicles is considered. The complete 3-D model of an antenna array and a UH-60 helicopter is constructed and simulated in a numerical electromagnetic tool, FEKO, and the array observations are obtained both in time and frequency. When the antenna arrays are mounted on such platforms, antenna pattern and characteristics change significantly leading to erroneous DF results. In this thesis, a new calibration technique is proposed when the vehicle is on the ground. In ground calibration, the major error sources are the reflections from the platform and the multipath from the ground. In order to mitigate these distortions, a time-gating method is proposed. When the received signals from the antennas are observed in time, the reflections are present with a time delay after the desired signal component. This part of the signals is gated and sufficiently clean calibration data is obtained for DF operation. The evaluation of the calibration data is done using both correlative interferometer and the MUSIC algorithms. The proposed method is advantageous for its simplicity, accuracy and cost effectiveness. In the second problem, a DF scenario is considered where the antenna array receives a mixture of a far-field signal and its near-field multipaths. Due to the model mismatch of the received signal components, both far- and near-field signal models should be used accordingly for accurate parameter estimation. Moreover, the array covariance matrix is rank-deficient due to multipath so that the subspace methods cannot be directly used for accurate DOA estimation. A new method is proposed for the estimation of DOA angles of far- and near-field signals and the ranges of near-field multipaths. 2-D DOA angle of the far-field source is estimated by using a calibration technique. A near-to-far-field transformation is proposed to suppress the far-field components of the array output so that the near-field source parameters can be estimated. Then spatial smoothing is employed to estimate the near-field source DOA angles. In order to estimate the near-field source ranges, a compressed sensing approach is presented where a dictionary with near-field sources with different ranges is employed. The proposed method is evaluated using close-to-real world data generated by a numerical electromagnetic tool, Wireless Insite, where the array and transmitter are placed in an irregular terrain and the array data is generated using full 3-D propagation model. It is shown that unknown source parameters can be estimated effectively showing the potential of the proposed approach in applications involving high-frequency direction finding and indoor localization. In the third problem, 2-D DOA and MC coefficient estimation is considered for arbitrary array structures. Previous methods in the literature are usually proposed for certain array geometries and show limited performance at low SNR or for small number of snapshots. In this thesis, compressed sensing is used to exploit the joint-sparsity of the array model to estimate both DOA and MC coefficients with a single snapshot for an unstructured array where the antennas are placed arbitrarily in space. A joint-sparse recovery algorithm for a single snapshot (JSR-SS) is presented by embedding the source DOA angles and MC coefficients into a joint-sparse vector. A dictionary matrix is defined by considering the symmetricity of the MC matrix for the unstructured antenna array. The proposed method is extended to the multiple snapshots, and the joint-sparse recovery algorithm with multiple snapshots (JSR-MS) is developed. A new joint-sparsity structure, namely, joint-block-sparsity is introduced to take advantage of the structure in the composite matrix involving both DOA and MC coefficients. In order to utilize the joint-block-sparsity effectively in the optimization problem, new norm structures, namely L-2,2,0- and L-2,2,1-norms are defined. The same technique is modified in order to solve the gain/phase mismatch problem in multipath scenario. Several simulations are done in order to show the performance of the proposed techniques.