Traffic Prediction of Space-Ground Integrated Information Network based on Improved LSTM Algorithm

A network traffic prediction system based on the modified LSTM algorithm was devised in response to the constraints of the standard BFS algorithm in space-ground integrated information network traffic prediction. Initial steps in developing a network traffic prediction scheme include precisely localizing influencing components using gating mechanism theory, properly dividing indicators to decrease interference, and finally, building the scheme using the upgraded LSTM algorithm. The experimental findings demonstrate that the suggested scheme outperforms the conventional BFS algorithm under certain assessment criteria, particularly with regard to the processing time of influencing variables and the accuracy of network traffic predictions. The ability to effectively forecast and optimize the development characteristics and product creation of the space-ground integrated information network is enabled by network traffic prediction, which plays a crucial role in this network. When it comes to tackling traffic simulation challenges, the classic BFS method has its limits, particularly when dealing with complicated situations. To address this issue more effectively, this research proposes a network traffic prediction technique that is based on an upgraded LSTM algorithm. This technique makes use of the upgraded LSTM algorithm to build itself after precisely locating the influencing components using the gating mechanism theory in order to decide the division of indicators. Under certain assessment conditions, the experimental findings reveal that the scheme's speed and accuracy are much enhanced for various tasks, and it performs better overall. Hence, for the purpose of improving the simulation accuracy and efficiency in space-ground integrated information network traffic prediction, a method based on the upgraded LSTM algorithm is preferable to the classic BFS algorithm.