Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Greedy and Transformer-Based Multi-Port Selection for Slow Fluid Antenna Multiple Access

Darian Perez-Adan, Jose P. Gonzalez-Coma, F. Javier Lopez-Martinez, Luis Castedo

Abstract

We address the port-selection problem in fluid antenna multiple access (FAMA) systems with multi-port fluid antenna (FA) receivers. Existing methods either achieve near-optimal spectral efficiency (SE) at prohibitive computational cost or sacrifice significant performance for lower complexity. We propose two complementary strategies: (i) GFwd+S, a greedy forward-selection method with swap refinement that consistently outperforms state-of-the-art reference schemes in terms of SE, and (ii) a Transformer-based neural network trained via imitation learning followed by a Reinforce policy-gradient stage, which approaches GFwd+S performance at lower computational cost.

Greedy and Transformer-Based Multi-Port Selection for Slow Fluid Antenna Multiple Access

Abstract

We address the port-selection problem in fluid antenna multiple access (FAMA) systems with multi-port fluid antenna (FA) receivers. Existing methods either achieve near-optimal spectral efficiency (SE) at prohibitive computational cost or sacrifice significant performance for lower complexity. We propose two complementary strategies: (i) GFwd+S, a greedy forward-selection method with swap refinement that consistently outperforms state-of-the-art reference schemes in terms of SE, and (ii) a Transformer-based neural network trained via imitation learning followed by a Reinforce policy-gradient stage, which approaches GFwd+S performance at lower computational cost.

Paper Structure

This paper contains 13 sections, 13 equations, 7 figures, 2 tables, 2 algorithms.

Table of Contents

  1. Introduction
  2. System Model
  3. Proposed Port Selection Methods
  4. GFwd with Swap Refinement
  5. Transformer-Based Neural Port Selection
  6. Input Features
  7. Architecture
  8. Two-Phase Training
  9. Numerical Results
  10. Computational Complexity Analysis
  11. SE Performance and Scalability Analysis
  12. Conclusions
  13. Non-Decreasing SINR Property of GFwd

Figures (7)

  • Figure 1: Transformer-based NN port selector architecture: 2 training phases.
  • Figure 2: Validation SE during training at SNR $\in \{10, 15, 20\}$ dB.
  • Figure 3: Average SE versus SNR for $P=100$, $L=8$, and $K=10$.
  • Figure 4: Average SE versus swap rounds $R$ for $P=100$, $L=8$, and $K=10$. Dashed lines indicate GEPort reference performance at each SNR.
  • Figure 5: Average SE vs. $K$ (users) for $P=100$, $L=8$, and $\mathrm{SNR}=15$ dB.
  • ...and 2 more figures