Applications of Adaptive Antenna for Wireless Communications
| Autor: | HSU,CHUN, 徐均 |
|---|---|
| Rok vydání: | 2008 |
| Druh dokumentu: | 學位論文 ; thesis |
| Popis: | 96 The smart antennas introduce the technique of antenna array into the field of wireless mobile communications which leads to the combination of antenna array and communication signal processing. One of the most important techniques for the smart antennas is the adaptive antenna array. For the adaptive array processing, adaptive beamforming and high-resolution direction-of-arrival (DOA) estimation are the two most important areas. Most of all the approaches for these two areas are based on the assumption of that the signal propagation model and antenna array characteristics are precisely known. But in practical situations, severe degradation is created by the impact of mismatches between the presumed and actual characteristics of the propagation medium and the receiving antenna, which also includes the array manifold mismodelling. Such mismatches could be induced by imperfect array calibration, distorted antenna shape, source local scattering, multipath propagation effects and environmental inhomogeneities. Under this circumstance, the DOA estimation techniques require to possess the robust capability against these mismatches, in which the robustness means it is insensitive to the perturbations of antenna model. For the adaptive beamformers, the requirement is the robustness and adaptive capability against the effect of insufficient data of reference signal and the pointing errors. First of all, under the environments of imperfect array with position perturbation and colored noise, this dissertation investigates the application of recursive algorithm to a forward linear predictor for DOA estimation of wireless communications. The proposed approach can deal with parameter’s uncertainty by minimizing the worst possible amplification of perturbations from imperfect array and noise signals. In comparison with the conventional algorithms for the DOA estimating performance, the proposed approach is consistently close to optimum from low to high signal-to-noise ratio (SNR) for static and moving source signals. Secondly, in the beamforming areas, this dissertation proposed an efficient technique for adaptive minimum variance distortionless response (MVDR) and eigenspace-based (ESB) beamforming with robust capabilities. To avoid the loss of degrees of freedom in suppressing undesired interferers, we proposed a maximum output magnitude searching technique to estimate the incident angle of the desired signal by using MVDR beamforming. Based on the calibrated result, the MVDR and ESB with the proposed robust technique can mitigate pointing error and product almost the same convergence speed as the MVDR and ESB beamformers under correct steering, respectively. For the application of conventional MVDR beamformer in dealing with the code division multiple access (CDMA) signals, an efficient derivative polynomial rooting calibration method that is robust in the scenarios of pointing errors, finite samples and multiple access interference (MAI) is proposed. Calibration process is done by estimating the pointing error of the desired signal. The main skill is to root the first-order derivative of the cost function instead of direct rooting it. In comparison with the conventional polynomial rooting method, the proposed approach can improve the root-selecting computation and reduced the bias of pointing error estimate due to the effect of noise and MAI. Finally, in concerning with the multipath environments created by sources local scatterings, this dissertation proposed an efficient subarray recursive least square (SRLS) beamformer with adaptive forgetting factor (AF) in which it can be adjusted in each adaptation process of the partitioned subweight. The proposed approach not only performs the adaptive beamforming, but also adaptively changes the forgetting factors to track the environmental changes. It is reliable in providing better performance than the conventional RLS-based algorithms but on the reduced computational burden basis. |
| Databáze: | Networked Digital Library of Theses & Dissertations |
| Externí odkaz: |
|