Beamforming for wireless communications

Abstract

This thesis studies the DOA estimation and beam pattern synthesis techniques for wireless communications applications. In particular, three specific problems have been investigated and novel algorithms have been proposed in this thesis. In wireless communications systems, the information about directions of signal of interests is very important for the subsequent signal processing procedures. Many modulated communication signals exhibit a cyclostationary property, which corresponds to the underlying periodicity arising from carrier frequencies or baud rates. Utilizing this underlying property in the received signals, we propose a new DOA (Direction of Arrival) estimation algorithm for cyclostationary signals. The proposed algorithm is very effective when the desired signals do not share a common cycle frequency. We have also derived the necessary conditions and provided a theoretical analysis on the choice of the parameters for making our proposed algorithm effective in coherent environment. Next, we come to beam pattern design techniques for antenna arrays. We classify the techniques into two categories: non-iterative beampattern synthesis algorithms and iterative beampattern synthesis algorithms. For non-iterative approaches, we propose an improved beampattern synthesis technique based on the Generalized Sidelobe Canceller (GSC) structure. Constrained optimization algorithm is used to control the beampattern from the main beam.