A Novel PAPR Reduction Scheme for OFDM System Based on Deep Learning

Kim, Minhoe; Lee, Woongsup; Cho, Dong-Ho

doi:10.1109/lcomm.2017.2787646

Cited by 200 publications

(142 citation statements)

References 10 publications

Supporting

Mentioning

136

Contrasting

Unclassified

Order By: Relevance

“…Training of the end-to-end system introduced in the previous section is performed w.r.t. to the loss function L defined in (6), as opposed to previous works which train autoencoders w.r.t. to the usual cross-entropy L defined in (5).…”

Section: Simulations Resultsmentioning

confidence: 99%

Joint Learning of Geometric and Probabilistic Constellation Shaping

Stark

Aoudia

Hoydis

2019

2019 IEEE Globecom Workshops (GC Wkshps)

View full text Add to dashboard Cite

The choice of constellations largely affects the performance of communication systems. When designing constellations, both the locations and probability of occurrence of the points can be optimized. These approaches are referred to as geometric and probabilistic shaping, respectively. Usually, the geometry of the constellation is fixed, e.g., quadrature amplitude modulation (QAM) is used. In such cases, the achievable information rate can still be improved by probabilistic shaping. In this work, we show how autoencoders can be leveraged to perform probabilistic shaping of constellations. We devise an information-theoretical description of autoencoders, which allows learning of capacity-achieving symbol distributions and constellations. Recently, machine learning techniques to perform geometric shaping were proposed. However, probabilistic shaping is more challenging as it requires the optimization of discrete distributions. Furthermore, the proposed method enables joint probabilistic and geometric shaping of constellations over any channel model. Simulation results show that the learned constellations achieve information rates very close to capacity on an additive white Gaussian noise (AWGN) channel and outperform existing approaches on both AWGN and fading channels.

show abstract

Section: Simulations Resultsmentioning

confidence: 99%

Joint Learning of Geometric and Probabilistic Constellation Shaping

Stark

Aoudia

Hoydis

2019

2019 IEEE Globecom Workshops (GC Wkshps)

View full text Add to dashboard Cite

show abstract

“…One naive approach for this issue includes the regularization of the cost function with a penalty parameter [18], [21], [22]. The corresponding unconstrained formulation can be expressed as…”

Section: B Conventional Training Methodsmentioning

confidence: 99%

“…The underlying idea is based on the primaldual method [31], which has been utilized for tackling traditional constrained optimization problems, to train the encoding and the decoding neural networks with multiple dimming constraints. We derive a single gradient decent optimization algorithm for the the DNN parameters and the dual variables, which is suitable for the existing DL optimization libraries to tackle the constrained training problem in (18). At the t-th iteration of the proposed training technique, the DNN parameters Θ D that minimizes (20) are computed by the steepest descent as…”

Section: Proposed Training Methodsmentioning

confidence: 99%

“…In particular, a machine learning framework called an autoencoder (AE) [23] has been employed to design a pair of wireless transmitter and receiver over fading channels via a single DL training process [16]. This result has been extended successfully to various scenarios such as orthogonal frequency division multiplexing (OFDM) [18] and optical fiber communication [19].…”

Section: A Related Work and Motivationmentioning

confidence: 99%

“…The recent works in [20]- [22] investigate the feasibility of the AE framework in the dimmable VLC system design. Compared to the RF counterparts in [16]- [18], the DLbased VLC optimization requires to control the behavior of DNNs by including optical signal constraints, e.g., intensity nonnegativity and dimming support, into the training process. A multi-colored VLC scenario is considered in [20] where red/green/blue LEDs are utilized for the message transmission.…”

Section: A Related Work and Motivationmentioning

confidence: 99%

See 2 more Smart Citations

A Deep Learning Approach to Universal Binary Visible Light Communication Transceiver

Lee

Quek

Lee

2020

IEEE Trans. Wireless Commun.

View full text Add to dashboard Cite

This paper studies a deep learning (DL) framework for the design of binary modulated visible light communication (VLC) transceiver with universal dimming support. The dimming control for the optical binary signal boils down to a combinatorial codebook design so that the average Hamming weight of binary codewords matches with arbitrary dimming target. An unsupervised DL technique is employed for obtaining a neural network to replace the encoder-decoder pair that recovers the message from the optically transmitted signal. In such a task, a novel stochastic binarization method is developed to generate the set of binary codewords from continuous-valued neural network outputs. For universal support of arbitrary dimming target, the DL-based VLC transceiver is trained with multiple dimming constraints, which turns out to be a constrained training optimization that is very challenging to handle with existing DL methods. We develop a new training algorithm that addresses the dimming constraints through a dual formulation of the optimization. Based on the developed algorithm, the resulting VLC transceiver can be optimized via the end-to-end training procedure. Numerical results verify that the proposed codebook outperforms theoretically best constant weight codebooks under various VLC setups.

show abstract

Peak to average power ratio alleviation by utilizing swarm intelligence along with machine learning for multiple input multiple output‐orthogonal frequency division multiplexing based underwater acoustic communication system

BanooriFarhad

ShiJinglun

ArshadJehangir

et al. 2021

Trans Emerging Tel Tech

View full text Add to dashboard Cite

An underwater acoustic (UWA) network uses multiple input multiple output‐orthogonal frequency division multiplexing (MIMO‐OFDM) to achieve high data rate with high amplitude that results in terms of augmented peak to average power ratio (PAPR). This significant PAPR deemed as a stumbling block to degrade the performance of MIMO‐OFDM in UWA communication by causing nonlinear distortion within the high‐power amplifier (HPA). While PAPR mitigation techniques have been suggested for the single input single output UWA case, there has been limited research work on controlling the PAPR dilemma in the MIMO‐OFDM UWA case. To address this flaw, this article proposes a modified behavior of the artificial bee colony algorithm at the transmission end, which is further cross‐verified using machine learning techniques. The simulation results show that the proposed technique outperforms current state‐of‐the‐art PAPR diminishing methods, which is further enhanced in accordance with varying neuron counts and population size. Following that, we also analyzed and compared energy efficiency and performance gain of our proposed technique along with various advanced techniques. Furthermore, a desirable ratio between new and trained data was obtained to improve network efficiency by keeping PAPR values low.

show abstract

A Novel PAPR Reduction Scheme for OFDM System Based on Deep Learning

Cited by 200 publications

References 10 publications

Joint Learning of Geometric and Probabilistic Constellation Shaping

Joint Learning of Geometric and Probabilistic Constellation Shaping

A Deep Learning Approach to Universal Binary Visible Light Communication Transceiver

Peak to average power ratio alleviation by utilizing swarm intelligence along with machine learning for multiple input multiple output‐orthogonal frequency division multiplexing based underwater acoustic communication system

Contact Info

Product

Resources

About