On Residual Energy Maximization in DF Cognitive Radio Networks With Multiple Eavesdroppers

This paper explores the issues of energy harvesting (EH) and multiple eavesdropping in an underlay cognitive radio network (CRN) that comprises of a secondary transmitter (SUT) and destination (SUD) node connected through a decode-forward (DF) relay (SR). CRN is operated following a frame structure that includes three non-overlapping slots where in the first one, the relay harvests energy from the information signal of SUTand the interference signal of primary user (PU). During the second slot, SUTtransmits data to the relay node. In the last slot, SRforwards the same to SUDand SUTdoes EH from PU signal. Residual energy maximization is done under the constraints of energy causality for SUTand SR, interference to PU, secondary outage probability and secrecy outage probability. Analytical expressions of ergodic secrecy capacity and secrecy outage probability are derived for non-colluding (NCE) and colluding eavesdropping (CE) scenarios. Closed form expressions for the optimal transmit power of SUTand SRalong with the time slot for EH are derived and their impacts on the residual energy, secondary outage probability and secrecy outage probability are analysed. Simulation results show a gain on the maximum value of the residual energy by ~8.13% for CE and ~4.72% for NCE over the existing works.