Skip to Main Content
Cognitive radio networks have the ability to efficiently utilize the scarce radio spectrum by allowing unlicensed users to access the licensed frequency bands in the absence of licensed users. Transmit beamformers can be designed by setting constraints on the interference temperature of the licensed users and signal to interference and noise ratios (SINRs) of the cognitive users. This design is however very sensitive to errors in channel state information (CSI). In this paper, we propose robust and non-robust beamforming techniques for multiuser cognitive radios. The proposed beamformer has the ability to maintain the SINRs of all unlicensed users above a target value for all possible errors in the CSI. The problem is formulated within a convex optimization framework with constraints on worst-case errors which can be solved using interior point methods. The performance of the robust beamformer is compared with non-robust beamformer in terms of BER of the unlicensed users and the probability density function of the achieved SINRs.