Low Complex Modulation Classification in NOMA Systems Using Weight Maximization Algorithm

In non-orthogonal multiple access (NOMA) systems, the modulation of the interfering users must be known for successive interference cancellation. Automatic modulation classification (AMC) techniques are employed in NOMA to reduce the signal processing overhead required to demodulate interfering signals. However, the existing feature-based approach is highly complex due to the covariance matrix computation in probability density function estimation. This letter presents a low-complexity feature-based approach to classify modulation schemes in three-user NOMA systems. We propose a weight maximization algorithm at the near user (NU) and intermediate user (IU) receivers, which effectively utilizes a weight factor computed using higher-order cumulants of the received superposed signal to classify the far user’s (FU) modulation scheme. Our algorithm achieved 95% classification accuracy at a signal to noise ratio (SNR) of 5 dB and 100% accuracy at a SNR of 12 dB for 800 symbols, with a power allocation factor of 8. Computational analysis showed a reduction of 89.5% in complex addition and 91.8% in complex multiplication operations compared to the state-of-the-art technique.