Transmit fading and interference mitigation for multi-antenna wireless communications

Wireless networks suffer from impairments that affect both the link performance and spectral efficiency. Two of the most significant impairments are microscopic fading and co-channel interference. In this thesis, we study transmitter-based schemes that mitigate both fading and interference.; The first part of the thesis presents a transmitter-based scheme to mitigate both fading to a desired receiver and co-channel interference to an undesired receiver (the interferee). We focus on downlink transmission with multiple antennas at the base station and single antennas at the intended and interferee receivers. Appropriate pre-coding and Space Time (ST) coding is combined to maximize link performance subject to a specified degree of co-channel interference suppression and assuming spatial correlation information at the transmitter. The solution to the fading and interference problem is a combination of a ST Block Code (STBC) and a multi-mode beamformer with mode directions and power allocation that depend on the level of the interference mitigation desired. The algorithm is able to smoothly trade-off the degrees-of-freedom provided by the transmit antennas between diversity and interference mitigation.; The second part of the thesis studies diversity and interference mitigation with an emphasis on the multi-cell interaction between transmit-receive pairs in separate cells. This interaction is measured by a utility metric derived from the error probability of each user. The interaction between the two users can be classified as greedy or co-operative. In the greedy algorithm, each user selfishly maximizes the diversity at their desired receiver using a combination of a multimode beamformer and a basic ST code. The users account for changes in the incoming interference as the other user optimizes its performance and they eventually reach an equilibrium called the Nash Equilibrium. In the co-operative algorithm, each user maximizes the diversity at its desired receiver but cooperates by limiting the amount of interference to the other user. We show that by using this algorithm, it is possible to find Pareto dominant beamformers where both users either maintain or improve their performance. Both schemes combine diversity coding with interference avoidance, one avoiding incoming interference at the desired receiver and the other avoiding out-going interference towards an interferee receiver.