Swarming drones deploy flying Ad-hoc networks to communicate with each other and also with ground stations,In some applications, each flying unit has to forward a large amount of traffic data, such as recorded video, images, and collected data. Besides, in this type of application, data traffic loss is very probable in wireless radio communications. Therefore, a more efficient technique is required to detect traffic loss and traffic retransmission. In this study, advanced time series prediction machine learning technique such as LSTM is applied to a Round Trip Time (RTT) series prediction. Result of different LSTM configuration and some tuning reached to the prediction accuracy of 95.6%.
展开▼