Optimal Power Allocation for Underlay-Based Cognitive Radio Networks with Primary User's Statistical Delay QoS Provisioning
Yichen Wang,Pinyi Ren,Qinghe Du,Li Sun
DOI: https://doi.org/10.1109/twc.2015.2462816
IF: 10.4
2015-01-01
IEEE Transactions on Wireless Communications
Abstract:Due to the highly-stochastic nature of wireless channels, how to provide efficient delay quality-of-service (QoS) provisioning for primary users (PU) while optimizing the performance of secondary users (SU) is a critically important task for cognitive radio networks (CRN). To address the above issue, we investigate the optimal power allocation strategy for underlay-based CRN with PU's statistical delay QoS protection. Instead of utilizing the widely-used interference power constraint to protect PU's transmission, we aim at satisfying PU's statistical delay QoS requirement characterized by the queue-length bound violation probability. By applying the theory of effective capacity, we further convert PU's queue-length bound violation probability constraint to the equivalent maximum sustainable traffic load requirement. Then, we formulate the optimization problem to maximize SU's average throughput while meeting PU's statistical delay QoS requirement as well as SU's average and peak transmit power constraints, which can be proved as a nonconvex problem. By employing the theories of convex hull and probabilistic transmission, we convert the original nonconvex problem to the equivalent strictly convex problem and then obtain the optimal power allocation strategy, which adapts to both PU's delay QoS requirements and channel conditions. Moreover, we also develop for comparison a fixed power allocation scheme that only adjusts with PU's delay QoS requirements. Simulation results are provided which demonstrate that both the optimal and fixed schemes can flexibly allocate the upperbounded transmit power budget according to PU's delay QoS requirements, but the proposed optimal power allocation strategy can also efficiently exploit the time-varying nature of wireless channels and thus significantly outperforms the fixed power allocation scheme.