Low Complexity Deep-Decoder for OTSM With Hardware Impairments

doi:10.1109/LCOMM.2023.3327249

Low Complexity Deep-Decoder for OTSM With Hardware Impairments

2023-12-01

4

PRE

AI

DOI：10.1109/LCOMM.2023.3327249

Original

Original request for help

Save

Abstract

En 中文

Recently, the single-carrier modulation scheme orthogonal time-sequency multiplexing (OTSM) has emerged as an alternative to orthogonal time-frequency space (OTFS). By virtue of using Walsh-Hadamard transform, OTSM reduces modulation/demodulation complexity, with a performance similar to OTFS. However, the bit error-rate (BER) performance of OTSM deteriorates drastically in presence of transceiver hardware impairments (HIs), particularly with in-phase and quardature phase imbalance. In this letter, we develop a deep neural network based signal detector for OTSM (called DL-OTSM) to effectively compensate for HIs, with low decoding complexity. We extensively evaluate the performance of DL-OTSM with respect to variations in system parameters, including user mobility, HIs, frame size, and modulation order. Our results show that DL-OTSM achieves a significantly better BER performance compared to the state-of-the-art Gauss-Seidel method and conventional minimum mean-square error detector, regardless of HIs.

Keywords:

Symbols

Signal detection

Hardware

Convolutional neural networks

Detectors

Transmitters

Delays

DL-OTSM

hardware-impairments (HIs)

IQ imbalance

signal detection

deep learning

data augmentation

Low Complexity Deep-Decoder for OTSM With Hardware Impairments

Abstract

Journal

Organization