A Residual Learning-Aided Convolutional Autoencoder for SCMA

Article Engineering, Electrical & Electronic

A Flexible Encoding/Decoding Procedure for 6G SCMA Wireless Networks via Adversarial Machine Learning Techniques

Luciano Miuccio et al.

Summary: 5G networks may not be fully capable of meeting the requirements of high data rates, system capacity, and spectral efficiency in the upcoming 6G networks. One potential solution is the adoption of Sparse Code Multiple Access (SCMA), a code-domain Non-Orthogonal Multiple Access (NOMA) technique. However, implementing SCMA in practical networks faces challenges in designing an optimal codebook set and an efficient decoding algorithm. This paper presents an end-to-end SCMA en/decoding structure that combines a state-of-the-art AutoEncoder (AE) architecture and a novel Wasserstein Generative Adversarial Network (WGAN) to address these challenges.

IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY (2023)

Add to Collection

Article Computer Science, Information Systems

A Wasserstein GAN Autoencoder for SCMA Networks

Luciano Miuccio et al.

Summary: In this study, an end-to-end SCMA encoding and decoding structure is designed based on a state-of-the-art autoencoder architecture and a novel Wasserstein Generative Adversarial Network (WGAN) that improves robustness to channel noise. The performance of the proposed network is compared with conventional computationally intensive solutions and a deep learning based autoencoder, and shows better Symbol Error Rate (SER) performance, especially at low energy per bit to noise power spectral density ratio.

IEEE WIRELESS COMMUNICATIONS LETTERS (2022)

Add to Collection

Article Computer Science, Information Systems

A Novel Multitask Learning Empowered Codebook Design for Downlink SCMA Networks

Qu Luo et al.

Summary: Sparse code multiple access (SCMA) is a promising non-orthogonal multiple access (NOMA) scheme for massive machine-type communication. This study leverages deep learning to jointly design the downlink SCMA encoder and decoder using an autoencoder, resulting in improved sparse codebooks and low-complexity decoder. Experimental results show significant improvements in error rate and computational complexity compared to conventional SCMA schemes.

IEEE WIRELESS COMMUNICATIONS LETTERS (2022)

Add to Collection

Article Computer Science, Information Systems

SCMA Decoding via Deep Learning

Chia-Po Wei et al.

Summary: SCMA has emerged as a competitive technology for future cellular systems. By utilizing a DNN method, the computational complexity of the decoder can be reduced, resulting in better BER performance and lower complexity compared to other DNN solutions previously studied.

IEEE WIRELESS COMMUNICATIONS LETTERS (2021)

Add to Collection

Article Telecommunications