Table of Contents
Fetching ...

Enhancing Sum-Rate Performance in Constrained Multicell Networks: A Low-Information Exchange Approach

Youjin Kim, Jonggyu Jang, Hyun Jong Yang

TL;DR

This work tackles sum-rate optimization in constrained multicell downlink networks with limited antenna counts and backhaul capacity. It introduces a DQN-based distributed beamforming approach that selects binary intercell weighting coefficients $\boldsymbol{\beta}_{i,t}$ under local CSI, reducing information exchange to $N_C$ bits per slot while adapting to time-varying channels. The method yields measurable per-cell rate gains over conventional schemes, with offline-trained policies enabling rapid online operation. Practically, the approach offers a scalable path toward higher spectral efficiency in realistic networks where global CSI and extensive inter-BS communication are impractical.

Abstract

Despite the extensive research on massive MIMO systems for 5G telecommunications and beyond, the reality is that many deployed base stations are equipped with a limited number of antennas rather than supporting massive MIMO configurations. Furthermore, while the cell-less network concept, which eliminates cell boundaries, is under investigation, practical deployments often grapple with significantly limited backhaul connection capacities between base stations. This letter explores techniques to maximize the sum-rate performance within the constraints of these more realistically equipped multicell networks. We propose an innovative approach that dramatically reduces the need for information exchange between base stations to a mere few bits, in stark contrast to conventional methods that require the exchange of hundreds of bits. Our proposed method not only addresses the limitations imposed by current network infrastructure but also showcases significantly improved performance under these constrained conditions.

Enhancing Sum-Rate Performance in Constrained Multicell Networks: A Low-Information Exchange Approach

TL;DR

This work tackles sum-rate optimization in constrained multicell downlink networks with limited antenna counts and backhaul capacity. It introduces a DQN-based distributed beamforming approach that selects binary intercell weighting coefficients under local CSI, reducing information exchange to bits per slot while adapting to time-varying channels. The method yields measurable per-cell rate gains over conventional schemes, with offline-trained policies enabling rapid online operation. Practically, the approach offers a scalable path toward higher spectral efficiency in realistic networks where global CSI and extensive inter-BS communication are impractical.

Abstract

Despite the extensive research on massive MIMO systems for 5G telecommunications and beyond, the reality is that many deployed base stations are equipped with a limited number of antennas rather than supporting massive MIMO configurations. Furthermore, while the cell-less network concept, which eliminates cell boundaries, is under investigation, practical deployments often grapple with significantly limited backhaul connection capacities between base stations. This letter explores techniques to maximize the sum-rate performance within the constraints of these more realistically equipped multicell networks. We propose an innovative approach that dramatically reduces the need for information exchange between base stations to a mere few bits, in stark contrast to conventional methods that require the exchange of hundreds of bits. Our proposed method not only addresses the limitations imposed by current network infrastructure but also showcases significantly improved performance under these constrained conditions.
Paper Structure (10 sections, 12 equations, 5 figures, 2 tables, 1 algorithm)

This paper contains 10 sections, 12 equations, 5 figures, 2 tables, 1 algorithm.

Figures (5)

  • Figure 1: Example of the state of the proposed DQN algorithm of BS 1 over the time slot $t$ for $N_C=3$
  • Figure 2: Example of the proposed DQN algorithm for $N_C=3$
  • Figure 3: Per-cell average rate versus time slot .
  • Figure 4: Per-cell average rate versus transmit power of BSs.
  • Figure 5: The required information exchange in bits.