Convolutional Neural Network and Long Short-Term Memory based Ensemble Learning for modulation classification in 5G Network
T Devi., Ahmad Alkhayyat, Vinaye Armoogum, B. Buvaneswari, B M Manjula · 2024
Modulation classification is important for its fundamental role in dynamic spectrum access which supports 5G (fifth generation) wireless communications. When the receiver receives multiple signals, multipath fading exists at the receiver side it is a big challenge to classify the modulation type. When many signals are transmitting to the receiver side, there exists a heterogeneous traffic, where there is a lot of disturbances occur. Due to these reasons, the block error rate for modulation increases simultaneously. To overcome the changing demands like multipath fading, minimize block error rate and to reduce the traffic. In this research a modulation classification algorithm is developed using the combination architectures of Convolutional Neural Network (CNN) and Long Short-Term Memory (LSTM) based Ensemble Learning (EL). Dimensionality is reduced through principal component analysis; it also minimizes the training time. The proposed method outperformed same as the ordinary CNN in terms of classification accuracy, but it reduced the training time with the help of Principal Component Analysis (PCA). The results proved that the proposed CNN-LSTM-EL achieved the best classification accuracy of 99.5 % than all other existing methods like CNN, LSTM.