A Robust Power Allocation Algorithm in Cognitive Radio Networks ⋆

Traditional distributed resources allocation algorithms are based on the hypothesis of perfect channel in cognitive radio networks. In fact, the parameters of the channel are variable, uncertain in some degree or even turbulent. Considering an ellipsoidal uncertainty set to describe the fluctuation of the interference gain, a robust power allocation algorithm is proposed to maximize the data rate of the cognitive users under both CR-to-PR interference constraint and the maximum transmit power for each cognitive user. In order to solve this optimization problem having infinite number of constraints, we transform it into an optimization problem with finite number of constraints under the worst case conditions. Computer experiments show that the proposed algorithm not only provides stability and robustness to the CR networks but also guarantee seamless communication.