Mathematical optimization and signal processing techniques for cooperative wireless networks

The rapid growth of mobile users and emergence of high data rate multimedia and interactive services have resulted in a shortage of the radio spectrum. Novel solutions are therefore required for future generations of wireless networks to enhance capacity and coverage. This thesis aims at addressing this issue through the design and analysis of signal processing algorithms. In particular various resource allocation and spatial diversity techniques have been proposed within the context of wireless peer-to-peer relays and coordinated base station (BS) processing. In order to enhance coverage while providing improvement in capacity, peer-to-peer relays that share the same frequency band have been considered and various techniques for designing relay coefficients and allocating powers optimally are proposed. Both one-way and two-way amplify and forward (AF) relays have been investigated. In order to maintain fairness, a signal-to-interference plus noise ratio (SINR) balancing criterion has been adopted. In order to improve the spectrum utilization further, the relays within the context of cognitive radio network are also considered. In this case, a cognitive peer-to-peer relay network is required to achieve SINR balancing while maintaining the interference leakage to primary receiver below a certain threshold. As the spatial diversity techniques in the form of multiple-input-multipleoutput (MIMO) systems have the potential to enhance capacity significantly, the above work has been extended to peer-to-peer MIMO relay networks. Transceiver and relay beamforming design based on minimum mean-square error (MSE) criterion has been proposed. Establishing uplink-downlink MSE duality, an alternating algorithm has been developed. A scenario where multiple users are served by both the BS and a MIMO relay is considered and a joint beamforming technique for the BS and the MIMO relay is proposed. With the motivation of optimising the transmission power at both the BS and the relay, an interference precoding design is presented that takes into account the knowledge of the interference caused by the relay to the users served by the BS. Recognizing joint beamformer design for multiple BSs has the ability to reduce interference in the network significantly, cooperative multi-cell beamforming design is proposed. The aim is to design multi-cell beamformers to maximize the minimum SINR of users subject to individual BS power constraints. In contrast to all works available in the literature that aimed at balancing SINR of all users in all cells to the same level, the SINRs of users in each cell is balanced and maximized at different values. This new technique takes advantage of the fact that BSs may have different available transmission powers and/or channel conditions for their users.