Robust resource allocation for orthogonal frequency division multiplexing-based cooperative cognitive radio networks with imperfect channel state information

In this study, the authors study the robust resource allocation problem for orthogonal frequency division multiplexing-based cooperative cognitive radio networks (CRNs) with decode and forward protocol and consideration of imperfect channel state information. The objective is to maximise the capacity of the cooperative CRN, while the interference to primary user receiver is below a predefined interference threshold and the transmit power of cognitive source and each relay is kept within their power budgets. Considering all possible channel uncertainties, they propose a heuristic robust relay selection scheme and formulate robust power allocation as a semi-infinite programming (SIP). By the worst-case approach, the SIP problem is converted into a convex optimisation problem and solved by the Lagrange dual decomposition method. They also analyse feasible regions of the constraints, convergence behaviour and computational complexity of their proposed robust algorithm. Simulation results show the impact of channel uncertainties and the outperformance of the proposed algorithm by comparing with non-robust algorithms.