Toward 6G: Deep GRU and RNN Empowered MVDR and LCMV Adaptive Beamformers for IRS-Aided Wireless Environments
6G systems require highly adaptive beamforming techniques to cope with rapid channel variations and strict latency constraints. The advent of Intelligent Reflecting Surfaces (IRS) has emerged as a transformative paradigm in the evolution of 6G wireless communication systems, enabling programmable ra...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/11072345/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1839625301055766528 |
|---|---|
| author | D. L. Sharini Ravilla Dilli M. Kanthi G. D. Goutham Simha |
| author_facet | D. L. Sharini Ravilla Dilli M. Kanthi G. D. Goutham Simha |
| author_sort | D. L. Sharini |
| collection | DOAJ |
| description | 6G systems require highly adaptive beamforming techniques to cope with rapid channel variations and strict latency constraints. The advent of Intelligent Reflecting Surfaces (IRS) has emerged as a transformative paradigm in the evolution of 6G wireless communication systems, enabling programmable radio environments through passive beam control. Driven by the potentials of IRS-assisted communication strategies, this work proposes a novel hybrid beamforming architecture for IRS-assisted Multiple-Input Multiple-Output (MIMO) systems. The architecture seamlessly integrates constrained optimization techniques—specifically Minimum Variance Distortionless Response (MVDR) and Linearly Constrained Minimum Variance (LCMV)—with advanced deep recurrent learning models, including Gated Recurrent Units (GRU) and Recurrent Neural Networks (RNN). The proposed approach enables precise signal alignment with improved spatial selectivity, effective interference suppression, and enhanced spectral efficiency. Monte Carlo simulations validate the efficient performance of the GRU-based MVDR beamforming scheme, achieving a spectral efficiency of ~5 bps/Hz observed at lower signal-to-noise ratios (SNRs), outperforming other hybrid approaches. Furthermore, Average Bit Error Probability (ABEP) analysis employing MVDR framework investigates the impact of transmit correlation and phase rotation in correlated Rayleigh fading scenarios. MVDR beamforming consistently boosts ABEP performance across all the transmit correlations by minimizing symbol misalignment during signal transmission. Additionally, the GRU-integrated MVDR approach achieves target BER levels at notably lower transmit power of ~8 dBm and enhancements of ~6 dBm under imperfect Channel State Information (CSI) scenarios in comparison with other schemes. |
| format | Article |
| id | doaj-art-7601ae7cd01c4f4080e38b2b716fb8c4 |
| institution | Matheson Library |
| issn | 2169-3536 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj-art-7601ae7cd01c4f4080e38b2b716fb8c42025-07-17T23:01:29ZengIEEEIEEE Access2169-35362025-01-011312061912063110.1109/ACCESS.2025.358672711072345Toward 6G: Deep GRU and RNN Empowered MVDR and LCMV Adaptive Beamformers for IRS-Aided Wireless EnvironmentsD. L. Sharini0https://orcid.org/0009-0007-1013-5290Ravilla Dilli1https://orcid.org/0000-0003-0450-8584M. Kanthi2https://orcid.org/0000-0002-5769-4272G. D. Goutham Simha3https://orcid.org/0000-0002-3908-3566Department of Electronics and Communication Engineering, Manipal Academy of Higher Education, Manipal, Manipal Institute of Technology, Karnataka, IndiaDepartment of Electronics and Communication Engineering, Manipal Academy of Higher Education, Manipal, Manipal Institute of Technology, Karnataka, IndiaDepartment of Electronics and Communication Engineering, Manipal Academy of Higher Education, Manipal, Manipal Institute of Technology, Karnataka, IndiaDepartment of Electronics and Communication Engineering, Manipal Academy of Higher Education, Manipal, Manipal Institute of Technology, Karnataka, India6G systems require highly adaptive beamforming techniques to cope with rapid channel variations and strict latency constraints. The advent of Intelligent Reflecting Surfaces (IRS) has emerged as a transformative paradigm in the evolution of 6G wireless communication systems, enabling programmable radio environments through passive beam control. Driven by the potentials of IRS-assisted communication strategies, this work proposes a novel hybrid beamforming architecture for IRS-assisted Multiple-Input Multiple-Output (MIMO) systems. The architecture seamlessly integrates constrained optimization techniques—specifically Minimum Variance Distortionless Response (MVDR) and Linearly Constrained Minimum Variance (LCMV)—with advanced deep recurrent learning models, including Gated Recurrent Units (GRU) and Recurrent Neural Networks (RNN). The proposed approach enables precise signal alignment with improved spatial selectivity, effective interference suppression, and enhanced spectral efficiency. Monte Carlo simulations validate the efficient performance of the GRU-based MVDR beamforming scheme, achieving a spectral efficiency of ~5 bps/Hz observed at lower signal-to-noise ratios (SNRs), outperforming other hybrid approaches. Furthermore, Average Bit Error Probability (ABEP) analysis employing MVDR framework investigates the impact of transmit correlation and phase rotation in correlated Rayleigh fading scenarios. MVDR beamforming consistently boosts ABEP performance across all the transmit correlations by minimizing symbol misalignment during signal transmission. Additionally, the GRU-integrated MVDR approach achieves target BER levels at notably lower transmit power of ~8 dBm and enhancements of ~6 dBm under imperfect Channel State Information (CSI) scenarios in comparison with other schemes.https://ieeexplore.ieee.org/document/11072345/Beamformingspectral efficiencyGRURNNdeep learningMVDR |
| spellingShingle | D. L. Sharini Ravilla Dilli M. Kanthi G. D. Goutham Simha Toward 6G: Deep GRU and RNN Empowered MVDR and LCMV Adaptive Beamformers for IRS-Aided Wireless Environments IEEE Access Beamforming spectral efficiency GRU RNN deep learning MVDR |
| title | Toward 6G: Deep GRU and RNN Empowered MVDR and LCMV Adaptive Beamformers for IRS-Aided Wireless Environments |
| title_full | Toward 6G: Deep GRU and RNN Empowered MVDR and LCMV Adaptive Beamformers for IRS-Aided Wireless Environments |
| title_fullStr | Toward 6G: Deep GRU and RNN Empowered MVDR and LCMV Adaptive Beamformers for IRS-Aided Wireless Environments |
| title_full_unstemmed | Toward 6G: Deep GRU and RNN Empowered MVDR and LCMV Adaptive Beamformers for IRS-Aided Wireless Environments |
| title_short | Toward 6G: Deep GRU and RNN Empowered MVDR and LCMV Adaptive Beamformers for IRS-Aided Wireless Environments |
| title_sort | toward 6g deep gru and rnn empowered mvdr and lcmv adaptive beamformers for irs aided wireless environments |
| topic | Beamforming spectral efficiency GRU RNN deep learning MVDR |
| url | https://ieeexplore.ieee.org/document/11072345/ |
| work_keys_str_mv | AT dlsharini toward6gdeepgruandrnnempoweredmvdrandlcmvadaptivebeamformersforirsaidedwirelessenvironments AT ravilladilli toward6gdeepgruandrnnempoweredmvdrandlcmvadaptivebeamformersforirsaidedwirelessenvironments AT mkanthi toward6gdeepgruandrnnempoweredmvdrandlcmvadaptivebeamformersforirsaidedwirelessenvironments AT gdgouthamsimha toward6gdeepgruandrnnempoweredmvdrandlcmvadaptivebeamformersforirsaidedwirelessenvironments |