Abstract
Cooperative beamforming or collaborative beamforming is an energy-efficient cooperative diversity technique that cophases multiple received signals at the receiver to compensate the phase differences due to defferent propagation delays. With perfect synchronization, a deterministic mainlobe power level can be guaranteed at the inteneded direction for a given network topology. However, the sidelobes, depending on the relays’ positions as well as the weighting vectors, have probabilistic power levels over the angle domain. Moreover, a high power level sidelobe may cause an unacceptable interference level at an unintended receiver. Thus, limiting the sidelobes to meet some requirements is of importance in cooperative beamforming. In this thesis, we develop a two-step multi-relay selection scheme for cooperative conventional beamforming based on a selection vector design. We formulate an optimization problem that minimizes the average power over all unintended directions with a constraint of unit power gain pointing to the desired user. Using the Lagrange multiplier, we further reduce the optimization problem as an eigenvalue decomposition problem. The solution is then used to form two selection criteria, where the first one is for finding the optimal reference relay and the second one is for selecting the optimal relaying set for cooperative conventional beamforming according to the reference relay. The proposed relay selection method provides a reduced-complexity, suboptimal solution for cooperative beamforming, where the receiver selects a set of cooperating relays by broadcasting a binary feedback message. Simulation results show that the proposed multi-relay selection scheme provides lower average power over all unintended directions than both the near and far cases, where the performance is close to that of the exhaustive search approach.