Abstract
Full-duplex (FD) enabled cognitive radio (CR) is promising since it can largely improve the spectrum efficiency by simultaneously learning the radio environments and delivering data traffic. The performance gain of FD CR is achieved under the premise that the self-interference (SI) can be suppressed to be sufficiently small. However, to what extent should the SI be suppressed so that FD CR is more advantageous than half-duplex (HD) CR? In addition, is SI the only factor that should be considered in leveraging HD CR and FD CR? The answers to these questions are investigated in this paper by studying degrees of freedom (DoF) for HD and FD CR. Theoretical analysis shows that not only SI but also the so-called time-antenna gain determines the DoF tradeoff between HD CR and FD CR. For all the CR schemes including interweave, underlay, overlay, and hybrid, the SI thresholds and normalized time-antenna gain with which FD CR can outperform HD CR are theoretically derived. Moreover, the DoF upper-bound of HD CR is inversely proportional to the primary traffic intensity, the primary interference power and the radio environments learning time. Nevertheless, the DoF upper-bound of FD CR only relates to the primary interference power and the CR antenna number.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call