Learning From Noisy Labels for MIMO Detection With One-Bit ADCs

Park, Jinsung; Lee, Namyoon; Hong, Song‐Nam; Jeon, Yo-Seb

doi:10.1109/lwc.2022.3230403

Cited by 6 publications

(4 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…[23] focused on low-resolution ADC massive MIMO systems, emphasizing achievable rates and bit allocation. study in [24] introduced a learning-based method for one-bit ADCs in massive MIMO systems. [25] addressed the challenges in massive MIMO systems concerning low-resolution DACs and proposed a 1-bit downlink precoding algorithm with Inherent complexity, NPhard problem.…”

Section: Quantization Techniques In Massive Mimo and Mmwave Systemsmentioning

confidence: 99%

Design and Optimization of Hybrid Precoders in Massive MIMO Systems: Leveraging Low-Resolution ADCs/DACs, Reconfigurable Intelligent Surfaces, and Deep Learning Algorithms

2024

IJIES

View full text Add to dashboard Cite

In this research, we propose an advanced hybrid precoding design for massive MIMO systems. Our approach integrates two innovative strategies: employing low-resolution ADCs and DACs with non-uniform quantization (NniQ) and implementing dynamic hybrid relay reconfigurable intelligent surfaces (DHRR-RIS). We utilize an iterative alternating minimization algorithm to improve spectral and energy efficiency in the first strategy. The second approach integrates DHRR-RIS with machine learning techniques, including adaptive back propagation neural network (ABPNN) for channel estimation, deep deterministic policy gradient (DDPG) algorithm for hybrid precoding and combining, fire hawk optimization (FHO) for DHRR-RIS, and enhanced fuzzy C-means (EFCM) for data clustering. These methodologies significantly enhance bit error rate (BER) and weighted sum rate (WSR) compared to traditional uniform quantization (UniQ) methods. Our results show that combining low-resolution ADCs/DACs with NniQ and DHRR-RIS, further optimized by machine learning, effectively reduces hardware complexity and power usage while markedly improving BER and WSR, offering a promising direction for efficient massive MIMO system development.

show abstract

Section: Quantization Techniques In Massive Mimo and Mmwave Systemsmentioning

confidence: 99%

Design and Optimization of Hybrid Precoders in Massive MIMO Systems: Leveraging Low-Resolution ADCs/DACs, Reconfigurable Intelligent Surfaces, and Deep Learning Algorithms

2024

IJIES

View full text Add to dashboard Cite

show abstract

“…Despite these efforts, the channel estimation error in lowresolution quantization based systems remains considerable compared to that in traditional high-resolution systems [12], [38]. This discrepancy becomes even more pronounced when the number of available pilot symbols is restricted [39].…”

Section: B Related Workmentioning

confidence: 99%

“…To achieve high data rates in quantized systems, the precise estimation of channel state information (CSI) plays a pivotal role [12]. However, obtaining accurate CSI poses a formidable challenge in such systems due to nonlinearities resulting from low-resolution quantization at the receiver, hindering the realization of the full benefits of low-resolution ADCs for efficient data communication [13].…”

Section: Introduction a Background And Motivationmentioning

confidence: 99%

The Cost of Noncoherence: Avoiding Channel Estimation Through Differential Encoding in Phase Quantized Systems

Gayan,

Inaltekin,

Senanayake

et al. 2023

IEEE Open J. Commun. Soc.

View full text Add to dashboard Cite

This paper proposes a novel approach that utilizes differential encoding to overcome the channel estimation problem in communication systems with low-resolution quantization receivers. For differentially encoded data, we derive the maximum likelihood detection rule for the canonical block-2 detectors, employing just two consecutive quantized observations at the channel output and without any receiver-side channel state information. We establish the optimality of this maximum likelihood detection rule within the class of block-L detectors, where L ≥ 3, under the condition that n = log 2 M , with n and M denoting the number of quantization bits and input alphabet size, respectively. The derived detector has a simple and easily implementable structure, comparing the quantization region indices of consecutive observations to determine the transmitted message index. By leveraging the structure of the derived optimum detector, we obtain the expression for the message error probability in Rayleigh fading wireless channels. Through asymptotic analysis in the high signal-to-noise ratio regime, we reveal a crucial finding that achieving the same diversity order as infinite bit quantization with full channel knowledge requires an additional two bits at the quantizer, in addition to the minimum requirement of log 2 M bits. One bit compensates for the low-resolution effect, while the other addresses the lack of channel knowledge. Finally, we conduct an extensive simulation study to demonstrate the performance of the optimum detectors and quantify the performance loss resulting from the absence of channel knowledge at the receiver.

show abstract

“…The authors of [7] investigated the massive machine-type communication with one-bit ADCs and presented an active user detection algorithm. In addition, the parameter learning algorithm was proposed for MIMO detection with 1-bit ADCs under constrained pilot length [8]. The authors of [9] solved the problem using a homotopy optimisation with a box relaxation.…”

mentioning

confidence: 99%

Learnable binary MIMO detection with negative penalty based on inexact ADMM

Kim,

Park

2024

Electronics Letters

View full text Add to dashboard Cite

In this letter, a new learnable binary MIMO detection algorithm in massive MIMO systems with one‐bit analog‐to‐digital converters is proposed. A negative penalty is introduced to transform a non‐convex constraint to a convex box constraint and a learnable iterative algorithm is presented based on inexact alternating direction methods of multipliers. The proposed MIMO detection method performs better than the existing binary MIMO detection methods even with lower computational complexity.

show abstract

Learning From Noisy Labels for MIMO Detection With One-Bit ADCs

Cited by 6 publications

References 14 publications

Design and Optimization of Hybrid Precoders in Massive MIMO Systems: Leveraging Low-Resolution ADCs/DACs, Reconfigurable Intelligent Surfaces, and Deep Learning Algorithms

Design and Optimization of Hybrid Precoders in Massive MIMO Systems: Leveraging Low-Resolution ADCs/DACs, Reconfigurable Intelligent Surfaces, and Deep Learning Algorithms

The Cost of Noncoherence: Avoiding Channel Estimation Through Differential Encoding in Phase Quantized Systems

Learnable binary MIMO detection with negative penalty based on inexact ADMM

Contact Info

Product

Resources

About