We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.IT

Change to browse by:

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Information Theory

Title: Long-Term CSI-based Design for RIS-Aided Multiuser MISO Systems Exploiting Deep Reinforcement Learning

Abstract: In this paper, we study the transmission design for reconfigurable intelligent surface (RIS)-aided multiuser communication networks. Different from most of the existing contributions, we consider long-term CSI-based transmission design, where both the beamforming vectors at the base station (BS) and the phase shifts at the RIS are designed based on long-term CSI, which can significantly reduce the channel estimation overhead. Due to the lack of explicit ergodic data rate expression, we propose a novel deep deterministic policy gradient (DDPG) based algorithm to solve the optimization problem, which was trained by using the channel vectors generated in an offline manner. Simulation results demonstrate that the achievable net throughput is higher than that achieved by the conventional instantaneous-CSI based scheme when taking the channel estimation overhead into account.
Comments: Under revision in IEEE journal. Keywords: Reconfigurable intelligent surface (RIS), intelligent reflecting surface (IRS)
Subjects: Information Theory (cs.IT); Signal Processing (eess.SP)
Cite as: arXiv:2111.12212 [cs.IT]
  (or arXiv:2111.12212v1 [cs.IT] for this version)

Submission history

From: Cunhua Pan [view email]
[v1] Wed, 24 Nov 2021 01:08:09 GMT (374kb,D)

Link back to: arXiv, form interface, contact.