Residual Stack-Aided Hybrid CNN-LSTM-Based Automatic Modulation Classification for Orthogonal Time-Frequency Space System
Anand Kumar, Manish, Udit Satija
Abstract
In this letter, for the first time, we propose an automatic modulation classification (AMC) method for orthogonal time-frequency space (OTFS) signal modulation using a hybrid convolutional neural network and long short-term memory (CNN-LSTM) network with a residual stack. The proposed method uses in-phase and quadrature-phase (IQ) of the received OTFS modulated signal to classify the received modulation accurately. We consider the six digital modulation schemes such as binary phase shift keying (BPSK), quadrature PSK (QPSK), minimum-shift keying (MSK), on-off keying (OOK), 4-amplitude shift keying (4ASK), and 8ASK for orthogonal time-frequency space (OTFS) in the delay-Doppler domain. Results depict that the proposed method achieves a high classification performance even at a low signal-to-noise ratio (SNR).