Deep Learning of Near Field Beam Focusing in Terahertz Wideband Massive MIMO Systems
A general non-uniform linear array that adopts PSs and TD units, where the user equipment is assumed to be within the near field region of the large-dimensional array.
Key ideas
- Leveraging a reinforcement learning based framework to learn the analog phase-shifter configurations in the non-uniform antenna array
- Proposing a signal model inspired critic network design that is able to significantly reduce the required power measurements
- Designing a low-complexity search algorithm to configure the TD units that help improve the wideband performance of the system
Applications
- Achieving beam focusing in wideband large antenna array systems that operate in the near field regime
- Learning reflection beams that are able to focus on users in the large reflecting surfaces aided near-field systems
- Enabling fast beam adaptation in the dynamic environments
More information about this research direction
Paper: Y. Zhang and A. Alkhateeb, “Deep Learning of Near Field Beam Focusing in Terahertz Wideband Massive MIMO Systems,” in IEEE Wireless Communications Letters, vol. 12, no. 3, pp. 535-539, March 2023, doi: 10.1109/LWC.2022.3233566.
Abstract: Employing large antenna arrays and utilizing large bandwidth have the potential of bringing very high data rates to future wireless communication systems. However, this brings the system into the near-field regime and also makes the conventional transceiver architectures suffer from the wideband effects. To address these problems, in this letter, we propose a low-complexity frequency-aware beamforming solution that is designed for hybrid time-delay and phase-shifter based RF architectures. To reduce the complexity, the joint design problem of the time delays and phase shifts is decomposed into two subproblems, where a signal model inspired online learning framework is proposed to learn the shifts of the quantized analog phase shifters, and a low-complexity geometry-assisted method is leveraged to configure the delay settings of the time-delay units. Simulation results highlight the efficacy of the proposed solution in achieving robust performance across a wide frequency range for large antenna array systems.
@ARTICLE{10004962,
author={Zhang, Yu and Alkhateeb, Ahmed},
journal={IEEE Wireless Communications Letters},
title={Deep Learning of Near Field Beam Focusing in Terahertz Wideband Massive MIMO Systems},
year={2023},
volume={12},
number={3},
pages={535-539},
doi={10.1109/LWC.2022.3233566}}
To reproduce the results in this paper:
Example: Steps to generate the results in this figure
- Download all the files of the GitHub repository given above.
- Run main.py in critic_net_training directory.
- After it is finished, there will be a file named critic_params_trsize_2000_epoch_500_3bit.mat that will be used in the next step.
- Run main.py in analog_beam_learning directory.
- After it is finished, run read_beams.py in the same directory.
- Copy the generated file, i.e., ULA_PS_only.mat to the td_searching directory.
- Run NFWB_BF_TTD_PS_hybrid_low_complexity_search_algorithm.m in Matlab, which will generate the figure shown right as result.
