Anzeige
Erschienen in:
2020 | OriginalPaper | Buchkapitel
Deep Learning Based Intelligent Congestion Control for Space Network
verfasst von : Kun Li, Huachun Zhou, Hongke Zhang, Zhe Tu, Guanglei Li
Erschienen in: Space Information Networks
Verlag: Springer Singapore
Aktivieren Sie unsere intelligente Suche, um passende Fachinhalte oder Patente zu finden.
Wählen Sie Textabschnitte aus um mit Künstlicher Intelligenz passenden Patente zu finden. powered by
Markieren Sie Textabschnitte, um KI-gestützt weitere passende Inhalte zu finden. powered by
Abstract
In order to alleviate the impact of network congestion on the spatial network running traditional contact graph routing (CGR) algorithm and DTN protocol, we propose a flow intelligent control method based on deep convolutional neural network (CNN). The method includes two stages of offline learning and online prediction to intelligently predict the traffic congestion trend of the spatial network. A CGR update mechanism is also proposed to intelligently update the CGR to select a better contact path and achieve a higher congestion avoidance rate. The proposed method is evaluated in the prototype system. The experimental results show that it is superior to the existing CGR algorithm in terms of transmission delay, receiver throughput and packet loss probability.