Deep Autoencoder Learning for Relay-Assisted Cooperative Communication Systems

Lu, Yuxin; Cheng, Peng; Chen, Zhuo; Li, Yonghui; Mow, Wai Ho; Vucetic, Branka

doi:10.1109/tcomm.2020.2998538

Cited by 24 publications

(19 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…In this section, we evaluate the proposed bit-wise and symbol-wise AE frameworks for the AF relay networks with practical SNR values. We utilize QPSK modulation similar to [14]. To train the proposed architectures we utilize SGD with Adam optimizer [32], where the weights of the dense layers are initialized with the Glorot initializer [33].…”

Section: Simulation Resultsmentioning

confidence: 99%

“…In particular, end-to-end learning-based relay networks using AE frameworks have been studied for AF relaying networks in [13], [18] and for DF relaying networks in [14]- [16], [19]. The authors in [18] studied a two-way AF relay network using a bit-wise AE performing only modulation design in 2-dimensions by employing NN-based multiple fully-connected (dense) layers at the AF relay node, while the authors in [13] studied a one-way AF relay network using a symbol-wise AE with NN-based multiple dense layers at the AF relay node.…”

Section: A Literature Reviewmentioning

confidence: 99%

“…Moreover, the removal of the CSI knowledge is studied in [13] for the AF relay network and in [14]- [16], [19] for DF relay networks. The authors in [13] employed a single NN at relay node to decode and re-encode the signal together, whereas the authors in [14]- [16], [19] employed a greater decoding power by explicitly decoding and encoding signals at the relay with the help of separate NN-based decoders and encoders. However, to the best of the authors' knowledge, none of the previous works have removed the necessity of the CSI knowledge without utilizing NN-based processing at the relay node either for a bit-wise or symbol-wise AE framework, and/or, have studied bit-wise AE-based differential coded-modulation designs for an AF relaying networks.…”

Section: A Literature Reviewmentioning

confidence: 99%

“…Furthermore, for an end-to-end learning framework without the CSI knowledge, the radio transformer network (RTN) was first introduced in [1], and since then has been widely employed for decoding the signal in the absence of CSI knowledge at the decoder, such as in [14], [19]. However, to the best of the authors' knowledge, none of the previous works have proposed an RTN for AF relay networks for either the symbol-wise or bit-wise AE frameworks.…”

Section: A Literature Reviewmentioning

confidence: 99%

“…For greater insights, we compare the CE loss functions from an informationtheoretic perspective and the estimated MI to analyze the convergence of the proposed AE frameworks. • We remove the need for CSI knowledge and noise variances of the links for the proposed AE-based frameworks, even without the NN-based processing at the relay node, unlike [14]- [16], [19]. We show that training the encoder and decoder in an end-to-end manner leads to block-by-block differential coded-modulation (demodulation) design in 2n-dimensional space, thereby enabling the AE framework to decode the signals without the CSI knowledge.…”

Section: B Contributionsmentioning

confidence: 99%

See 4 more Smart Citations

End-to-End Learning-Based Framework for Amplify-and-Forward Relay Networks

Gupta

Sellathurai

2021

IEEE Access

View full text Add to dashboard Cite

We study end-to-end learning-based frameworks for amplify-and-forward (AF) relay networks, with and without the channel state information (CSI) knowledge. The designed framework resembles an autoencoder (AE) where all the components of the neural network (NN)-based source and destination nodes are optimized together in an end-to-end manner, and the signal transmission takes place with an AF relay node. Unlike the literature that employs an NN-based relay node with full CSI knowledge, we consider a conventional relay node that only amplifies the received signal using CSI gains. Without the CSI knowledge, we employ power normalization-based amplification that normalizes the transmission power of each block of symbols. We propose and compare symbol-wise and bit-wise AE frameworks by minimizing categorical and binary cross-entropy loss that maximizes the symbol-wise and bit-wise mutual information (MI), respectively. We determine the estimated MI and examine the convergence of both AE frameworks with signal-to-noise ratio (SNR). For both these AE frameworks, we design coded modulation and differential coded modulation, depending upon the availability of CSI at the destination node, that obtains symbols in 2n-dimensions, where n is the block length. To explain the properties of the 2n-dimensional designs, we utilize various metrics like minimum Euclidean distance, normalized second-order and fourth-order moments, and constellation figures of merit. We show that both these AE frameworks obtain similar spherical coded-modulation designs in 2n-dimensions, and bit-wise AE that inherently obtains the optimal bit-labeling outperforms symbol-wise AE (with faster convergence under low SNR) and the conventional AF relay network with a considerable SNR margin.INDEX TERMS Amplify-and-forward, autoencoder, feed-forward neural networks, learning, and relay networks.

show abstract

Section: Simulation Resultsmentioning

confidence: 99%

Section: A Literature Reviewmentioning

confidence: 99%

Section: A Literature Reviewmentioning

confidence: 99%

Section: A Literature Reviewmentioning

confidence: 99%

Section: B Contributionsmentioning

confidence: 99%

See 3 more Smart Citations

End-to-End Learning-Based Framework for Amplify-and-Forward Relay Networks

Gupta

Sellathurai

2021

IEEE Access

View full text Add to dashboard Cite

show abstract

Deep SCMA receiver: A low‐complexity joint decoder and channel estimator for SCMA over time‐varying channels using RNNs

Gamal

Elsayed

Mehana

2023

Trans Emerging Tel Tech

View full text Add to dashboard Cite

Sparse code multiple access (SCMA) is a promising non‐orthogonal multiple access scheme for cellular Internet of things (IoT) due to its ability to support massive connectivity, grant‐free transmission and scalability. Inspired by the recent developments of deep learning for physical layer communications, we present a design of an uplink SCMA receiver using deep learning. We propose the use of recurrent neural networks (RNNs) for joint channel estimation and multiuser data detection of uplink SCMA under time‐varying Rayleigh channel using a single deep learning structure. The use of RNNs enables the receiver to learn the time correlation between the received samples with a very low pilot density and with low complexity. Compared with the conventional SCMA receiver, the simulation results show that the proposed deep learning‐based receiver can achieve BER performance similar to that of the conventional SCMA receivers (such as sparse pilot channel estimator and message‐passing algorithm detector) with very low pilot density and with much lower complexity. Moreover, the proposed SCMA receiver shows good resilience to small changes in the receiver speed (second‐order channel statistics) which enables the proposed deep learning receiver to work over a reasonable range of receiver speeds without parameter tuning. Fine tuning the network parameters to capture the variations of the channel, during online transmission using small data sets and small training period, is also checked and provides additional BER benefits.

show abstract