Adaptive MIMO Detector Based on Hypernetwork: Design, Simulation, and Experimental Test

Zhang, Jing; Wen, Chao-Kai; Jin, Shi

doi:10.1109/jsac.2021.3126064

Cited by 17 publications

(7 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Hypernetwork can increase the adaptability of the main network. Avoid retraining the main network when the internal parameters or the number of network layers need to be changed, which greatly reduces the training cost [19], [22]. In addition, there are some hypernetwork solutions that can improve the performance of the main network [25].…”

Section: Hypernetworkmentioning

confidence: 99%

“…Hypernetwork, first proposed in [19], uses one network to generate internal parameters for another network to reduce training costs and improve the flexibility of the layers of the main neural network [20]. To increase the adaptability of machine learning (ML)-based multiple input multiple output (MIMO) detection to different channel environments and noise levels, the authors of [21] and [22] introduced hypernetwork into ML-based MIMO detection and proposed HyperMMNet and HyperEPNet, respectively. The authors in [23] proposed the HyperRNN architecture for an end-to-end downlink channel estimation scheme for massive MIMO frequency division duplex (FDD) systems, achieving lower normalized mean square error (NMSE) for channel estimation and higher sum-rate for beamforming.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Hypernetwork Based Model-Driven Channel Neural Decoding

Liang,

Lam,

et al. 2024

IEEE Access

View full text Add to dashboard Cite

Channel decoding algorithms based on model-driven deep learning, also known as channel neural decoding algorithms, have received a lot of attention in recent years. However, the internal parameters and number of layers of the current channel neural decoding algorithm cannot be changed after training. Once changed, retraining of the channel neural decoding network is required. Hypernetwork is a neural network that can generate internal parameters for the main neural network to reduce the training cost of the main neural network and improve the flexibility of the main neural network. In this study, a novel hypernetwork based channel neural decoder is proposed for neural belief propagation algorithms (NBP), including the neural normalized min-sum (NNMS) and neural offset min-sum (NOMS) algorithm. According to the type of information interaction between the hypernetwork and the main decoding network, hypernetwork-based channel neural decoders can be divided into two types: static and dynamic. The internal parameters of the static hypernetwork-based channel neural decoder can be updated as needed without retraining of the main network. In addition to this benefit, the number of layers of the dynamic hypernetwork-based channel neural decoder can also be adjusted. Experimental results show that, compared with the existing NNMS decoding algorithms, the proposed hypernetwork-based NNMS decoding algorithms can achieve better performance on both low-density parity-check (LDPC) and Bose-Chaudhuri-Hocquenghem (BCH) codes.

show abstract

Section: Hypernetworkmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Hypernetwork Based Model-Driven Channel Neural Decoding

Liang,

Lam,

et al. 2024

IEEE Access

View full text Add to dashboard Cite

show abstract

“…Similarly, the unfolded version of the expectation propagation detector was proposed wherein damping factors are learned using meta-learning [142]. This detector was also extended using hypernetworks to achieve generalization to new channel realizations and noise levels but for typical values of many other system parameters [143]. The major drawback here is that DNN must be retrained for each set of new system parameters.…”

Section: Data Detection and Classificationmentioning

confidence: 99%

Achievable Rate of Near-Field Communications Based on Physically Consistent Models

Akrout

Shyianov

Bellili

et al. 2023

IEEE Trans. Wireless Commun.

View full text Add to dashboard Cite

Data-driven machine learning (ML) is promoted as one potential technology to be used in next-generations wireless systems. This led to a large body of research work that applies ML techniques to solve problems in different layers of the wireless transmission link. However, most of these applications rely on supervised learning which assumes that the source (training) and target (test) data are independent and identically distributed (i.i.d). This assumption is often violated in the real world due to domain or distribution shifts between the source and the target data. Thus, it is important to ensure that these algorithms generalize to out-of-distribution (OOD) data. In this context, domain generalization (DG) tackles the OODrelated issues by learning models on different and distinct source domains/datasets with generalization capabilities to unseen new domains without additional finetuning. Motivated by the importance of DG requirements for wireless applications, we present a comprehensive overview of the recent developments in DG and the different sources of domain shift. We also summarize the existing DG methods and review their applications in selected wireless communication problems, and conclude with insights and open questions.

show abstract

“…The importance of artificial intelligence (AI) and ML in wireless communications is growing quickly and already an integral part of 5G Release 18 [22]. AI and ML are expected to play and even more central role in 6G systems [2] and research is heavily focusing on ML-based baseband processing, ranging from atomistic (separate and independent) optimizations of neural network (NN)-based MIMO data detectors [23]- [25] or channel decoders [26], to fully NN-based transmitters and receivers (e.g., end-to-end learning methods [27]- [30]), and model-driven deep unfolding approaches [20], [21], [31]. However, existing NN-based data detectors turn out to require orders of magnitude higher complexity compared to classical algorithms that were designed by hand.…”

Section: B Related Workmentioning

confidence: 99%

“…For the same reason, deep unfolding was utilized to optimize either classical data detection or channel decoding algorithms by hyperparameter tuning, or augment classical algorithms with a hypernetwork. Both methods were applied in [31] for improving a classical expectation propagation (EP) data detector in a deep-unfolded IDD receiver. Other examples for training a traditional LDPC decoder include message damping and weighted belief-propagation decoding [20], [21].…”

Section: B Related Workmentioning

confidence: 99%

DUIDD: Deep-Unfolded Interleaved Detection and Decoding for MIMO Wireless Systems

Wiesmayr

Dick

Hoydis

et al. 2022

2022 56th Asilomar Conference on Signals, Systems, and Computers

View full text Add to dashboard Cite

Iterative detection and decoding (IDD) is known to achieve near-capacity performance in multi-antenna wireless systems. We propose deep-unfolded interleaved detection and decoding (DUIDD), a new paradigm that reduces the complexity of IDD while achieving even lower error rates. DUIDD interleaves the inner stages of the data detector and channel decoder, which expedites convergence and reduces complexity. Furthermore, DUIDD applies deep unfolding to automatically optimize algorithmic hyperparameters, soft-information exchange, message damping, and state forwarding. We demonstrate the efficacy of DUIDD using NVIDIA's Sionna link-level simulator in a 5G-near multiuser MIMO-OFDM wireless system with a novel low-complexity soft-input soft-output data detector, an optimized low-density parity-check decoder, and channel vectors from a commercial raytracer. Our results show that DUIDD outperforms classical IDD both in terms of block error rate and computational complexity.

show abstract

Adaptive MIMO Detector Based on Hypernetwork: Design, Simulation, and Experimental Test

Cited by 17 publications

References 32 publications

Hypernetwork Based Model-Driven Channel Neural Decoding

Hypernetwork Based Model-Driven Channel Neural Decoding

Achievable Rate of Near-Field Communications Based on Physically Consistent Models

DUIDD: Deep-Unfolded Interleaved Detection and Decoding for MIMO Wireless Systems

Contact Info

Product

Resources

About