Experimental data set

To validate the proposed channel prediction algorithm, a 5G-TSN channel model was simulated in this study to develop a channel delay data model under experimental scenarios. Channel delay datasets were collected at different times through a sampling method. Using the collected dataset, the 5G-TSN mobile communication channel prediction method based on DELM-ELM-AE was developed, and the DELM model was optimized using the DBO algorithm. The prediction performance of the channel and the actual network operation indicators were then simulated and compared for validation.

This study compares the DELM model optimized with the DBO algorithm to the original CSI without the use of intelligent algorithms after collecting the channel delay dataset under a 5G-TSN laboratory scenario. The impact of neural networks and intelligent optimization algorithms on the 5G-TSN channel prediction performance was analyzed. Additionally, the number of neurons and time steps of the DELM-ELM-AE model were varied to simulate MSE of the 5G-TSN channel prediction.