Power saving for real-time services in multiuser OFDMA-based cognitive radio systems under average interference constraint

This paper considers a power minimization problem for Quality-of-Service provisioning in the downlink of multiuser OFDMA-based cognitive radio systems with real-time services by jointly optimizing over subcarrier and power allocation. An average interference constraint is imposed to protect the primary transmission. The optimal solution is obtained by using Lagrange methods. It can be seen that the optimal power allocation follows a modified water-filling approach with different water levels for different subcarriers and users. An optimal algorithm and a suboptimal algorithm based on stochastic nature are then proposed. Numerical comparisons show that the performance of the suboptimal algorithm with low complexity is close to that of the optimal one, which demonstrate that the suboptimal algorithm outperforms other algorithms.