Two-stage deep learning-based hybrid precoder design for very large scale massive MIMO systems

Jeyakumar, P.; Ramesh, Arvind; Srinitha, S.; Nishant, V.T.; Gowri, P.; Muthuchidambaranathan, P.

doi:10.1016/j.phycom.2022.101835

Cited by 3 publications

(3 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Extending the work in [17] on BFNN, the output dimensions of the analog beamforming (BF) matrix have been increased from Nt to NRF × Nt. This expansion allows for a higher level of flexibility and granularity in the BF matrix, providing more control and optimization capabilities.…”

Section: Major Contributionsmentioning

confidence: 99%

“…In [17][18][19] to achieve robustness against imperfect channel state information (CSI), a 2-stage design approach is suggested for the construction of the Beamforming Neural Network (BFNN). In the initial offline stage of training, in order to acquire an approaching methodology for near-optimal or perfect spectral efficiency (SE), BFNN is trained utilizing realistic channel estimates as input.The BFNN may therefore dynamically adapt and alter itself to account Optimal RF-Chain Selection with Hybrid Analog and Digital Beamformer Design Usin for imprecise CSI, achieving robustness against error in the channel estimate, in the subsequent stage of online deployment.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Optimal RF-Chain Selection with Hybrid Analog and Digital Beamformer Design Using Deep Learning in mmWave Communication System

Kotikalapudi,

Kamj

et al. 2023

Preprint

View full text Add to dashboard Cite

Energy efficiency is treated as an important performance metric in wireless communication systems, and selecting the appropriate RF chains plays a vital role in maximizing overall system performance, energy efficiency , and spectral efficiency. To maximize energy efficiency and spectral efficiency, a two-level optimization technique for RF chain selection and beamforming is proposed in this paper. In the first level, RF chain selection is performed using particle swarm optimization (PSO) algorithm to maximize system rate. The selected RF chains are then used in the second level, where a two-stage deep learning-based Beamform-ing Neural Network (BFNN) is employed to optimize the beamforming process while being robust to imperfect channel state information (CSI). The BFNN takes the estimated CSI as input and outputs the optimized analog beamformer. The proposed BFNN is trained using a loss function tailored for beamforming optimization. Experimental results validate the effectiveness of the proposed method, showcasing significant improvements in spectral efficiency, while mitigating the effects of imperfect CSI. The spectral efficiency with and without RF selection using PSO is 51.5439 and 47.7258 respectively at SNR of 20dB. This approach provides an efficient and robust solution for RF chain selection and beamforming optimization in wireless communication systems.

show abstract

Section: Major Contributionsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Optimal RF-Chain Selection with Hybrid Analog and Digital Beamformer Design Using Deep Learning in mmWave Communication System

Kotikalapudi,

Kamj

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…DL techniques, including a learned denoising-based approximate message passing (LDAMP) network and a spatial-frequency convolutional neural network (SF-CNN), were required to reconstruct millimeter-wave channels. Consideration is given to the relationships between space and frequency by these methods [12,13].…”

Section: Introductionmentioning

confidence: 99%

Metaheuristic hybrid deep learning-based channel estimation and precoding for millimeter-wave massive MIMO system

Shailender,

Khera,

Singh

2024

Preprint

View full text Add to dashboard Cite

In mm-wave massive MIMO systems, the inefficiency of the digital or analogue precoders necessitates that each antenna has its own radio frequency chain. From this standpoint, a method of hybrid precoding that is economically viable has been devised. Digital beamformers are used for controlling antenna elements using short-dimensional precoding. From this standpoint, a method of hybrid precoding that is economically viable has been devised. During the channel estimation phase, we propose using deep recurrent convolutional neural networks (Deep RCNNs). Salp Swarm Algorithm (SSA) and Tuna Swarm Optimization (TSO) algorithms are combined in a hybrid meta-heuristic algorithm to improve Deep RCNN's channel estimation efficiency. O-BiLSTM, or optimised bidirectional long short-term memory, is utilised by hybrid precoding in order to execute the optimal precode across the anticipated channel. In relation to spectral efficiency and normalised mean square error (NMSE), the enhanced channel estimation approach surpasses the existing schemes through the implementation of Deep RCNN. The spectrum efficacy of the O-BiLSTM hybrid precoder design method is observed to be superior when compared to alternative approaches.

show abstract

Deep Learning Framework for Spectral Efficient Intelligent Hybrid Beamforming

Yousuf,

Sambhavi,

Nalband

et al. 2023

2023 International Conference on Recent Advances in Electrical, Electronics &Amp; Digital Healthcare Technologies (REEDCON)

View full text Add to dashboard Cite

Two-stage deep learning-based hybrid precoder design for very large scale massive MIMO systems

Cited by 3 publications

References 19 publications

Optimal RF-Chain Selection with Hybrid Analog and Digital Beamformer Design Using Deep Learning in mmWave Communication System

Optimal RF-Chain Selection with Hybrid Analog and Digital Beamformer Design Using Deep Learning in mmWave Communication System

Metaheuristic hybrid deep learning-based channel estimation and precoding for millimeter-wave massive MIMO system

Deep Learning Framework for Spectral Efficient Intelligent Hybrid Beamforming

Contact Info

Product

Resources

About