Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Antenna Coding Optimization for Pixel Antenna Empowered Wireless Communication Using Deep Learning with Heterogeneous Multi-Head Selection

Binzhou Zuo, Shanpu Shen, Hongyu Li

TL;DR

This work proposes a novel deep learning-based antenna coding optimization algorithm supported by a heterogeneous multi-head selection mechanism, whose main idea is to train multiple neural networks based on various coding schemes and select the one that leads to the best system performance.

Abstract

Pixel antenna is a promising antenna technology that enables flexible adjustment of radiation characteristics and enhancement of wireless systems through antenna coding. This work proposes a novel deep learning-based antenna coding optimization algorithm. Specifically, the proposed algorithm is supported by a heterogeneous multi-head selection mechanism, whose main idea is to train multiple neural networks based on various coding schemes and select the one that leads to the best system performance. Unlike traditional heuristic searching-based algorithms that require high computational complexity to achieve satisfactory performance, the proposed data-driven deep learning approach can achieve 98\% of the performance achieved by the searching-based algorithms with significantly reduced computational complexity. Results demonstrate that in pixel antenna empowered single-input single-output systems, the proposed algorithm achieves a computational speed 81 times faster than the searching-based algorithm. For more complex pixel antenna empowered multiple-input multiple-output systems, the computational speed is 297 times faster than the existing searching-based algorithm. Benefiting from the high performance and low computational complexity, this algorithm demonstrates the significant potential of pixel antennas as a novel and practical technology to enhance wireless systems.

Antenna Coding Optimization for Pixel Antenna Empowered Wireless Communication Using Deep Learning with Heterogeneous Multi-Head Selection

TL;DR

This work proposes a novel deep learning-based antenna coding optimization algorithm supported by a heterogeneous multi-head selection mechanism, whose main idea is to train multiple neural networks based on various coding schemes and select the one that leads to the best system performance.

Abstract

Pixel antenna is a promising antenna technology that enables flexible adjustment of radiation characteristics and enhancement of wireless systems through antenna coding. This work proposes a novel deep learning-based antenna coding optimization algorithm. Specifically, the proposed algorithm is supported by a heterogeneous multi-head selection mechanism, whose main idea is to train multiple neural networks based on various coding schemes and select the one that leads to the best system performance. Unlike traditional heuristic searching-based algorithms that require high computational complexity to achieve satisfactory performance, the proposed data-driven deep learning approach can achieve 98\% of the performance achieved by the searching-based algorithms with significantly reduced computational complexity. Results demonstrate that in pixel antenna empowered single-input single-output systems, the proposed algorithm achieves a computational speed 81 times faster than the searching-based algorithm. For more complex pixel antenna empowered multiple-input multiple-output systems, the computational speed is 297 times faster than the existing searching-based algorithm. Benefiting from the high performance and low computational complexity, this algorithm demonstrates the significant potential of pixel antennas as a novel and practical technology to enhance wireless systems.
Paper Structure (12 sections, 11 equations, 5 figures)

This paper contains 12 sections, 11 equations, 5 figures.

Table of Contents

  1. Introduction
  2. Antenna Coding Based on Pixel Antenna
  3. Pixel Antenna Model
  4. Formulation for Antenna Coding Optimization
  5. Deep Learning-Based Antenna Coding Design
  6. Architecture of HMSM
  7. Three-Feature Maps Input
  8. Heterogeneous Multi-Head Module
  9. Unzip Layer and Decision Function
  10. Training Process
  11. Performance Evaluation
  12. Conclusion

Figures (5)

  • Figure 1: (a) An example of pixel antenna designs proposed in 3zhang2022low and (b) the multi-port circuit network for pixel antenna.
  • Figure 2: Diagram of (a) pixel antenna empowered SISO systems and (b) pixel antenna empowered MIMO systems integrated with HMSM.
  • Figure 3: (a) Architecture of deep learning-based antenna coding design algorithm. (b) The major components for the model training process, which include the dataset preparation process and the training process. (c) An example of encoding pixel antenna with $Q$ pixel ports into two antenna maps.
  • Figure 4: Evaluation of the regression performance of the deep learning model for (a) binary code, (b) Gray code, (c) performance improvement of HMSM, and (d) learning accuracy versus number of samples in the dataset.
  • Figure 5: Performance comparison of different algorithms under (a) SISO and (b) MIMO systems, and (c) computational time consumption comparison between the HMSM algorithm and searching-based algorithms.