Graph-Theory-Based Resource Allocation and Mode Selection in D2D Communication Systems: The Role of Full-Duplex

We propose graph-theory-based mode selection and resource allocation algorithms for device-to-device (D2D) communication systems in which a full-duplex (FD) scheme can be applied. To apply FD in practice, the current half-duplex (HD) communication devices need to be upgraded to FD devices. Replacing all the devices at once, though, is undesirable; implementing FD incurs a high cost. It is prudent then to select the devices that ought to be changed first. This would ensure better performance at reducing frequency range or enhancing sum-rate. To resolve this issue, we suggest an algorithm that selects D2D pairs based on the centrality concept of graph theory. We also propose a resource allocation algorithm modified for FD-applied environments to reduce the range of frequency bands. Numerical analyses confirm that the proposed system achieves a reduction of the used frequency range as well as the theoretical maximum sum-rate performaperformancence.