Robust transmit power control with imperfect csi using a deep neural network

Woongsup Lee, Kisong Lee

Research output: Contribution to journalArticlepeer-review


In this paper, a robust transmit power control scheme is proposed for multi-channel underlay device-to-device (D2D) communications with imperfect channel state information (CSI). The transmit power of the D2D user equipment (DUE) on each channel is optimized to maximize the average spectral efficiency (SE) whilst maintaining the quality-of-service (QoS) of the cellular user equipment (CUE) in the presence of errors in the CSI. To this end, we propose a novel deep neural network (DNN) structure and training methodology, in which artificially distorted CSI is used to compensate for the effect of imperfect CSI, such that a robust transmit power control strategy against channel error can be derived. Our simulation results show that even when the CSI is inaccurate, in our proposed scheme the degradation of the average SE can be kept low whilst maintaining negligible QoS violation, thereby confirming its effectiveness and robustness.

Original languageEnglish
Pages (from-to)12266-12271
Number of pages6
JournalIEEE Transactions on Vehicular Technology
Issue number11
Publication statusPublished - 2021 Nov 1

Bibliographical note

Publisher Copyright:
© 2021 Institute of Electrical and Electronics Engineers Inc.. All rights reserved.

All Science Journal Classification (ASJC) codes

  • Automotive Engineering
  • Aerospace Engineering
  • Electrical and Electronic Engineering
  • Applied Mathematics


Dive into the research topics of 'Robust transmit power control with imperfect csi using a deep neural network'. Together they form a unique fingerprint.

Cite this