Transmit beamforming in OFDM systems and cognitive radio networks

Abstract

Transmit beamforming is a technique used in MIMO systems to increase diversity in the communication. We consider a broadcast scenario (downlink) with several antennas at the transmitter and one antenna at each receiver. In this masterthesis we propose techniques that determine power allocation and beamformers in an OFDM scenario. In Chapter 2 we make the scenario description and present some previous works in the single channel case. We describe two optimal transmit beamforming algorithms, and present one solution for the cognitive radio case. In Chapter 3 we introduce the problematic of the OFDM optimization and justify the motivation behind it. We explain the interest in developing algorithms that allocate resources in a femtocell and consider the cognitive radio case. Within this scope we formulate two different approaches to our problem: one considers power minimization while having specific constraints on every subcarrier of each user, and the second minimizes power while considering only one constraint per user. In Chapter 4 we develop two algorithms that suboptimally solve the problem with specific constraints on every subcarrier. For feedback signaling reasons, the users are imposed to use a single beamformer for all their subcarriers. Again, one of these solutions is extended to the cognitive radio case. In Chapter 5 we propose a new algorithm that converges towards a feasible solution in the case we use Exponential Effective SINR Mapping (EESM) constraints on every user. We also discuss its optimality and extend it again to the cognitive case. Finally, in Chapter 6 we present some simulations and results for these algorithms.