AbstractsComputer Science

Cooperative beamforming for cognitive radio (CR) systems, which uses geographically distributed transmitters to perform beamforming, can improve spectrum utilization. However, transmissions from geographically distributed cooperative transmitters introduce asynchronous interferences at the primary receivers (PRs) and the secondary receivers (SRs) due to the propagation delays. In this thesis, we develop data rate fairness cooperative beamforming techniques for a CR system with multiple SRs and multiple PRs in presence of asynchronous interferences. In particular, the optimal beamforming design is formulated as an optimization problem to maximize the minimum data rate of the SRs subject to transmission power constraints of the secondary cooperative transmitters (SCTs) as well as asynchronous interference power constraints at the PRs. The optimal beamforming problem is a quadratically constrained quadratic program (QCQP) max-min optimization problem, which is non-convex and non-linear. Therefore, we reformulate the optimal beamforming problem as a quasi-convex problem using the semidefinite program (SDP) relaxation that can be solved using the standard SDP solvers and bisection method. We study important properties of the optimization problem. By exploiting these properties, we also develop low complexity suboptimal beamforming techniques. Further, we extend the beamforming techniques to incorporate the uncertainties in the channel and propagation delay estimation between the SCTs and the PRs. We present numerical results in order to compare the complexity, the data rate fairness performance and the robustness of the developed beamforming techniques. Presented numerical results show that the developed suboptimal beamforming techniques provide trade-offs between data rate fairness performance and computational complexity. Moreover, the developed optimal and suboptimal beamforming techniques always guarantee the target interference threshold at the PRs.