Energy efficient design for non-orthogonal AF relaying in underlay spectrum sharing networks

In this paper, we address the power allocation problem in an underlay cognitive network that employs multiple non-orthogonal amplify-and-forward relays. The power allocation at the relaying nodes is designed to maximize the energy efficiency of the secondary network. We adopt the power-normalized signal-to-noise ratio (PN-SNR) as an efficiency measure. In this context, we formulate an optimization problem to design the power allocation adhering to an interference constraint forced by the primary network. Moreover, each relay has a maximum allowable transmission power to prolong the lifetime of the relaying nodes. Using a series of transformations, the problem is shown to be a product of two quadratic fractional functions, which is non-convex. We propose an algorithm to solve the problem using a transformation into a 1-D search problem. Simulation results show that the proposed power allocation is more efficient than the SNR-maximizing allocation and the equal power allocation and barely affects the outage probability of the system.