Hybrid Neural Coded Modulation: Design and Training Methods
Not on the map. Why?
Abstract (arXiv)
We propose a hybrid coded modulation scheme which composes of inner and outer codes. The outer-code can be any standard binary linear code with efficient soft decoding capability (e.g. low-density parity-check (LDPC) codes). The inner code is designed using a deep neural network (DNN) which takes the channel coded bits and outputs modulated symbols. For training the DNN, we propose to use a loss function that is inspired by the generalized mutual information. The resulting constellations are shown to outperform the conventional quadrature amplitude modulation (QAM) based coding scheme for modulation order 16 and 64 with 5G standard LDPC codes.
Key facts. Cited by 2 papers in the map (data through 7 Oct 2026). Builds on 18 of its 32 references that are in the map. First posted on arXiv 4 Feb 2022.
Builds on
18 of 32 references.
References · 18
- 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,160
- 2015Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shiftcited by 6,979
- 2017An Introduction to Deep Learning for the Physical Layercited by 227
- 2017Deep Learning-Based Communication Over the Aircited by 72
- 2017Deep Learning Methods for Improved Decoding of Linear Codescited by 66
- 2018Model-Driven Deep Learning for Physical Layer Communicationscited by 50
- 2019Turbo Autoencoder: Deep learning based channel codes for point-to-point communication channelscited by 42
- 2018Deepcode: Feedback Codes via Deep Learningcited by 41
- 2018Model-free Training of End-to-end Communication Systemscited by 29
- 2019Trainable Communication Systems: Concepts and Prototypecited by 22
- 2021Autoregressive Belief Propagation for Decoding Block Codescited by 15
- 2019Joint Learning of Geometric and Probabilistic Constellation Shapingcited by 13
6 more
- 2019Meta-Learning to Communicate: Fast End-to-End Training for Fading Channelscited by 13
- 2018LEARN Codes: Inventing Low-latency Codes via Recurrent Neural Networkscited by 9
- 2019"Machine LLRning": Learning to Softly Demodulatecited by 7
- 2019Autoencoder-Based Error Correction Coding for One-Bit Quantizationcited by 6
- 2019Neural Turbo Equalization: Deep Learning for Fiber-Optic Nonlinearity Compensationcited by 4
- 2019End-to-end Learning for GMI Optimized Geometric Constellation Shapecited by 3
Built on by
Citing papers per year: 2022: 1; 2023: 1.
2 papers.