On Residual Energy Maximization in Cognitive Relay Networks With Eavesdropping

Abstract

This paper addresses a residual energy maximization problem on a two-hop decode-and-forward relay-assisted cognitive radio network operating in a time-slotted mode in the presence of an eavesdropper. In the first time slot, spectrum sensing (SS) and energy harvesting (EH) from the received signal of the primary user (PU) are done simultaneously by the secondary transmitter through power splitting mode. On the next time slot, the secondary transmitter either harvests energy or transmits its own data to the relay node based on the SS decision result. The relay node subsequently either forwards this data or harvests energy based on the activity followed by the secondary transmitter node. Considering both linear and EH models, an optimization nonlinear problem of residual energy maximization is explored under the constraints of sensing reliability of the PU, secondary outage probability, and secrecy outage probability. Closed-form expressions of the sensing time and secondary and relay transmit power are derived both in the presence and absence of an eavesdropper. The impact of eavesdropping on detection probability, false alarm probability, secondary outage probability, and residual energy are analyzed. Simulation results show that about $\sim$ 21.19% and $\sim$ 42.84% gain in total residual energy values are achieved compared to the existing works.