Robust precoding for orthogonal space-time block coded MIMO cognitive radio networks

We consider a spectrum sharing based cognitive radio wireless communications system which consists of a secondary transmitter (ST) and a secondary receiver (SR), both equipped with multiple antennas, and a single-antenna primary user (PU). ST is composed of an orthogonal space-time block code (OSTBC) followed by a linear precoder which acts as a multimode beamformer. Given a fixed OSTBC, our goal is to design a robust linear precoder, where the robustness assumes addressing the problem of imperfect channel state information (CSI) of ST to SR (ST-SR) MIMO link at ST. Channel correlation information of ST to PU (ST-PU) link is also available at ST. The imperfect channel knowledge of the ST-SR link is described by a specified uncertainty region, known to ST. The robust precoding design criterion is the maximization of the signal-to-noise ratio (SNR) for the worst-case Frobeneus norm-bounded CSI mismatch subject to the transmit-power constraint of ST and interference-power constraint of PU. The original robust precoder design problem is non-convex and does not have any close-form solution. Using a semidefinite relaxation (SDR) approach, we transform it to a convex form and find the approximate solution using standard numerical methods. Simulation results are presented to show the performance of the robust precoding scheme.