Abstract:This article proposes the use of RNNs for joint channel estimation and multiuser data detection of uplink SCMA under time‐varying Rayleigh channel using a single deep learning structure. Compared with the conventional SCMA receiver, the proposed deep learning‐based receiver can achieve BER performance similar to that of the conventional SCMA receivers 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 which enables the proposed deep learning receiver to work over a reasonable range of receiver speeds without parameter tuning. 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.

Neural Network-Assisted Receiver Design Via Learning Trellis Diagram Online

Online Learning of Trellis Diagram Using Neural Network for Robust Detection and Decoding

Neural Network-Aided BCJR Algorithm for Joint Symbol Detection and Channel Decoding

End-to-End Learning for Uplink MU-SIMO Joint Transmitter and Non-Coherent Receiver Design in Fading Channels.

Symbol-Level Online Channel Tracking for Deep Receivers

A Deep Learning Based Receiver for Wireless Communications Systems with Unknown Channel Models.

Data-Driven Symbol Detection for Intersymbol Interference Channels with Bursty Impulsive Noise

Neural Network-Assisted Robust Symbol Detection Under Intersymbol Interference.

Design of Noncoherent Communications: From Statistical Method to Machine Learning

Collaborative Learning for Less Online Retraining of Neural Receivers

OFDM Receiver Design With Learning-Driven Automatic Modulation Recognition

Adaptive and Flexible Model-Based AI for Deep Receivers in Dynamic Channels

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

Meta-ViterbiNet: Online Meta-Learned Viterbi Equalization for Non-Stationary Channels

Generative-Adversarial-Network Enabled Signal Detection for Communication Systems with Unknown Channel Models

Joint Model and Data-Driven Receiver Design for Data-Dependent Superimposed Training Scheme with Imperfect Hardware

On the Robustness of Deep Learning-aided Symbol Detectors to Varying Conditions and Imperfect Channel Knowledge

Receiver Design for ICI-CSK System: A New Perspective Based on GRU Neural Network

Neural Network-based OFDM Receiver for Resource Constrained IoT Devices

Bayesian Receiver Design for Asynchronous Massive Connectivity

Aider: Artificial Intelligent Based Deep Receiver for Wireless Communication Systems