网络智能传输研究进展

发布时间:2023-10-24 作者:廖乙鑫,王子逸,崔勇

 

摘要:新兴超低时延场景的出现以及6G技术与人工智能技术的发展,使网络智能传输成为研究热点。分析了传输层和应用层的时延组成及影响因素,对机器学习技术与传输层、应用层流媒体传输相结合的智能传输协议的发展和优缺点进行了综述。从传统网络传输协议的发展、人工智能技术的发展、网络传输和人工智能结合3个方面展望了网络智能传输面临的机遇与挑战。认为分布式机器学习训练场景的传输性能、训练数据的质量、模型的泛化能力、模型大规模部署的开销是未来网络智能传输技术的重点研究方向。

关键词:网络传输;机器学习;时延分析;拥塞控制;视频传输

 

Abstract: With the emergence of emerging ultra-low latency scenarios and the development of 6G technology and artificial intelligence technology, intelligent network transmission has become a research hotspot. The delay components and influencing factors of the transport layer and application layer are discussed. Then, the development, advantages, and disadvantages of intelligent transmission protocols that combine machine learning technology with transport layer and application layer streaming media transmission are reviewed. The opportunities and challenges faced by intelligent network transmission are prospected from three aspects: the development of traditional network transmission protocols, the development of artificial intelligence technology, and the combination of network transmission and artificial intelligence. It is believed that the transmission performance of distributed machine learning training, the quality of training data, the generalization ability of the model, and the cost of large-scale deployment of the model is the key research direction of future network intelligent transmission technology.

Keywords: network transmission; machine learning; analysis of delay; congestion control; video transmission

在线PDF浏览: PDF