DeepRx: Fully Convolutional Deep Learning Receiver

Honkala, Mikko; Korpi, Dani; Huttunen, Janne M. J.

doi:10.1109/twc.2021.3054520

Cited by 96 publications

(76 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A natural choice for this would be a CNN. In recent years, CNNbased architectures have shown great results for the physical layer (see [4], [10]), and these use ResNet blocks [11]. As corroborated by simulation results in Section IV, ResNetbased demappers are performance-limited for our task since the correlation between users is not captured.…”

Section: Convolutional Attention-based Neural Demappermentioning

confidence: 56%

“…More recent techniques propose to jointly replace a group of functional blocks by DNNs. An example is [4], which proposes to jointly learn channel estimation, equalization, and LLR generation (demapping) using a DNN. There have also been several proposals to perform end-to-end learning -the joint optimization of the transceiver using an auto-encoder [5].…”

Section: Introductionmentioning

confidence: 99%

“…In many instances, it might not be feasible to have a very complex neural network (NN) in a product due to the large number of trainable parameters. For example, the NN of [4] has around 1.2 million parameters for uplink MIMO reception and detection, and this scales with the number of users. So, it might be more practical in some cases to use a simple channel estimator like the linear minimum mean square error (LMMSE) channel estimator at pilot locations, and employ a This work was carried out when A. Michon and F. Ait Aoudia were at Nokia Bell Labs France.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Convolutional Self-Attention-Based Multi-User MIMO Demapper

Athur¹,

Aoudia²,

Srinath³

2022

Preprint

View full text Add to dashboard Cite

In orthogonal frequency division multiplexing (OFDM)-based wireless communication systems, the bit error rate (BER) performance is heavily dependent on the accuracy of channel estimation. It is important for a good channel estimator to be capable of handling the changes in the wireless channel conditions that occur due to the mobility of the users. In recent years, the focus has been on developing complex neural network (NN)based channel estimators that enable an error performance close to that of a genie-aided channel estimator. This work considers the other alternative which is to have a simple channel estimator but a more complex NN-based demapper for the generation of soft information for each transmitted bit. In particular, the problem of reversing the adverse effects of an imperfect channel estimator is addressed, and a convolutional self-attention-based neural demapper that significantly outperforms the baseline is proposed.

show abstract

Section: Convolutional Attention-based Neural Demappermentioning

confidence: 56%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Convolutional Self-Attention-Based Multi-User MIMO Demapper

Athur¹,

Aoudia²,

Srinath³

2022

Preprint

View full text Add to dashboard Cite

show abstract

“…For example, a 2×2 MIMO scenario with a simple comb pilot structure is used for channel estimation in Ref. [57]. After interpolation, a three-layer fully connected neural network is used for optimal fitting.…”

Section: Intelligent Wireless Transmission Model Design 51 Channel Estimationmentioning

confidence: 99%

An intelligent wireless transmission toward 6G

Zhang

Niu

et al. 2021

Intell. and Converged Netw.

View full text Add to dashboard Cite

“…visual tracking [34] and receiver design [35]. To the authors' best knowledge, it has not yet been applied to RIS configuration.…”

Section: Introductionmentioning

confidence: 99%

Reconfigurable Intelligent Surface Enabled Spatial Multiplexing with Fully Convolutional Network

Peng¹,

Termöhlen²,

Sun³

et al. 2022

Preprint

View full text Add to dashboard Cite

Reconfigurable intelligent surface (RIS) is an emerging technology for future wireless communication systems. In this work, we consider downlink spatial multiplexing enabled by the RIS for weighted sum-rate (WSR) maximization. In the literature, most solutions use alternating gradient-based optimization, which has moderate performance, high complexity, and limited scalability. We propose to apply a fully convolutional network (FCN) to solve this problem, which was originally designed for semantic segmentation of images. The rectangular shape of the RIS and the spatial correlation of channels with adjacent RIS antennas due to the short distance between them encourage us to apply it for the RIS configuration. We design a set of channel features that includes both cascaded channels via the RIS and the direct channel. In the base station (BS), the differentiable minimum mean squared error (MMSE) precoder is used for pretraining and the weighted minimum mean squared error (WMMSE) precoder is then applied for fine-tuning, which is nondifferentiable, more complex, but achieves a better performance. Evaluation results show that the proposed solution has higher performance and allows for a faster evaluation than the baselines. Hence it scales better to a large number of antennas, advancing the RIS one step closer to practical deployment.

show abstract

DeepRx: Fully Convolutional Deep Learning Receiver

Cited by 96 publications

References 26 publications

Convolutional Self-Attention-Based Multi-User MIMO Demapper

Convolutional Self-Attention-Based Multi-User MIMO Demapper

An intelligent wireless transmission toward 6G

Reconfigurable Intelligent Surface Enabled Spatial Multiplexing with Fully Convolutional Network

Contact Info

Product

Resources

About