Dynamic Programming Based Power Control Algorithm with Primary User QoS Guarantee for Cognitive Radio Networks

To maximize the average rate in a longterm scope while guarantee primary user's QoS demand is a new key issue in cognitive radio systems. A dynamic programming based power control algorithm with the consideration of primary user's QoS by a rate loss constraint criterion is proposed in this paper. In the proposed algorithm, the occupancy for each subcarrier by primary users is modeled as a discrete-time Markov chain. And the concept of rate loss constraint is defined which is used to guarantee primary user's desired rate. Then the dynamic programming framework with rate loss constraint of this problem is formulated, which is a function of the primary user occupancy state, the system channel gains and the remaining power budgets for the cognitive radio transmitters. At last, we give a solution for the problem. Simulation results show that the proposed algorithm can obtain a maximum average data rate over a finite time horizon and effectively guarantee primary users' QoS.