A Novel Low-Complexity Precoding Algorithm for MIMO Cognitive Radio Systems

This paper proposes a novel efficient precoding algorithm to maximize the rate of the cognitive link, where a pair of multiple-input multiple-output cognitive radio users shares the spectrum allocated to multiple primary users as long as the interference power is acceptable. In our method, the precoding vectors are firstly obtained by solving an unconstrained convex problem which is formulated by using the penalty function idea and has a closed-form solution. Then, power is allocated to different precoding vectors to satisfy both the transmit-power constraint and interference-power constraints through interior method. Moreover, an effective set of the penalty weights is presented. Our method has much lower complexity than the optimal solution and similar complexity to other low-complexity methods. Simulation results show that our method significantly outperforms the existing low-complexity methods and has almost the same performance as the optimal solution.