Optimal Sensing and Power Control for Cognitive Radio Networks

We consider a cognitive radio network where secondary cognitive users communicate with an access point relying on opportunistic transmissions over wireless fading channels. For such a network, it is important to maximize aggregate utility of average user rates under the constraints of average user power budgets and maximum allowable probabilities of collisions with the primary communications. To this end, jointly optimal sensing selection and power control scheme for the cognitive users is derived in a quasi-closed form through a Lagrange dual-based gradient approach. Moreover, we develop a stochastic optimization algorithm that can operate without a-priori knowledge of the fading channel statistics. It is established that the proposed stochastic scheme is capable of approaching the optimal sensing and power control strategy.