Study on robust narrowband beamformers

• Main Author: Li, Xiaohui.
• Other Authors: Ser Wee
• Format: Final Year Project
• Language: English
• Published: 2009
• Subjects: DRNTU::Engineering
• Online Access: http://hdl.handle.net/10356/17998
• Institution: Nanyang Technological University

In the recent years, the number of wireless applications has increased drastically. The ever growing demand for the wireless communications services and other applications are constantly increasing the need for better coverage, enhancing the channel capacity and higher quality service. Using beamforming techniques, the quality of signal of interest can be enhanced, and the same carrier frequency can be reused to increase the channel capacity in the multi-user environment. In view of these demands, the objective of this project is to study some existing robust beamforming techniques using computer simulation and develop beamforming software with GUI to display the results obtained. From practical approach, this report looks into the robustness of the beamformers, namely: Standard Capon Beamformer, Robust Capon Beamformer and Iterative Robust Capon Beamformer. The beamforming algorithms are implemented using MATLAB to evaluate and compare their performances in numerical experiments. A new Adaptive Iterative Robust Capon Beamfomer algorithm has been proposed. Moving on to the larger scale, the next step is then to apply these beamforming techniques with the user defined simulation data and pre-recorded experimental data, thereby creating better comparison of the results. To provide a better presentation of the proposed beamfomer in comparison against other robust techniques, a beamforming software including the Graphic User Interface has been developed. From the empirical results, the proposed beamforming technique has shown significant improvement in convergence rate than the Iterative Robust Capon Beamforming (refer to Figure 6-7). In addition, the stopping criterion is no longer critical as it always converges to the desired steering vector. This is supported by the theoretical analysis and the computer simulations illustrate the effectiveness of the proposed method.