Optimal and sub-optimal resource allocation in multi-hop cognitive radio networks with primary user outage constraint

In this study, the authors propose a cross-layer optimisation framework for cognitive radio networks (CRNs) by jointly taking into account the physical layer and transport layer. In particular, they focus on joint rate and power control for a multi-hop CRN to maximise the network utility without affecting the performance of primary users (PUs). The formulation is shown to be a non-linear non-convex optimisation problem. To solve the problem, the authors first use a successive convex approximation method, which has been proved to converge to the global optimum. Then, the authors propose a novel heuristic method to develop a practical sub-optimal distributed algorithm without explicit message passing. The authors have proved that the proposed distributed algorithm is able to converge to the unique fixed point near the global optimum. Finally, the illustrative results indicate that the proposed sub-optimal algorithm outperforms the high-signal-to-interference-ratio-based algorithm and closely approaches the global optimum.