Downlink Power Control: A Performance Booster for Full Duplex Cellular Networks

<div><div>Full duplex (FD) communication uses the same time-frequency resource elements for uplink and downlink transmissions. Compared to half-duplex (HD) systems, this improves the effective spectrum usage, but at the cost of additional self-interference and increased co-channel interference (CCI). Promising improvements in full duplex downlink performance are well established in the literature, but the improvement in uplink is marginal or even severely degraded due to strong down-link signals. This paper focuses on managing CCI in an FD network by spatially controlling downlink and uplink transmit powers. We show that uplink performance can be improved through the introduction of downlink power control, but this slightly reduces downlink performance. We introduce a distance-based user scheduling to address this reduction. Furthermore, we have also shown that optimal overall performance can be achieved by designing appropriate uplink and downlink power control factors. We used the Matern Cluster process to model the network. The analytical results we formulated is verified by extensive Monte Carlo simulations. This work is submitted to IEEE transaction on Wireless Communications. Currently it is under review.</div></div>

Downlink Power Control: A Performance Booster for Full Duplex Cellular Networks

<div><div>Full duplex (FD) communication uses the same time-frequency resource elements for uplink and downlink transmissions. Compared to half-duplex (HD) systems, this improves the effective spectrum usage, but at the cost of additional self-interference and increased co-channel interference (CCI). Promising improvements in full duplex downlink performance are well established in the literature, but the improvement in uplink is marginal or even severely degraded due to strong down-link signals. This paper focuses on managing CCI in an FD network by spatially controlling downlink and uplink transmit powers. We show that uplink performance can be improved through the introduction of downlink power control, but this slightly reduces downlink performance. We introduce a distance-based user scheduling to address this reduction. Furthermore, we have also shown that optimal overall performance can be achieved by designing appropriate uplink and downlink power control factors. We used the Matern Cluster process to model the network. The analytical results we formulated is verified by extensive Monte Carlo simulations. This work is submitted to IEEE transaction on Wireless Communications. Currently it is under review.</div></div>