基于WebRTC的低延迟流媒体传输方法研究  

Research on Low Delay Streaming Media Transmission Method Based on WebRTC

作　　者：陈震 卢金勤 朱奇 朱磊 CHEN Zhen;LU Jin-qin;ZHU Qi;ZHU Lei(Jiangxi Kingroad Technologies Development Co.,Ltd.,Nanchang Jiangxi 330025,China)

机构地区：[1]江西锦路科技开发有限公司,江西南昌330025

出　　处：《机电产品开发与创新》2024年第6期29-31,共3页Development & Innovation of Machinery & Electrical Products

摘　　要：基于WebRTC的低延迟流媒体传输方法研究针对实时音视频通信中的延迟问题,通过集成深度强化学习(Deep Reinforcement Learning,DRL)优化视频速率控制,并与标准Google拥塞控制算法(Google Congestion Control,GCC)比较。研究在简单、中等和困难网络状态下评估模型性能,结果表明DRL模型在丢包率和往返时延上均优于GCC,尤其在困难模式下展现出更强的适应性和鲁棒性,为低延迟流媒体传输提供了有效解决方案。Research on low-latency streaming media transmission method based on WebRTC Aiming at the delay problem in real-time audio and video communication,the video rate control is optimized by integrating Deep Reinforcement Learning(DRL)and compared with the standard Google Congestion Control algorithm(GCC).The performance of the model is evaluated in simple,medium and difficult network conditions.The results show that DRL model is superior to GCC in packet loss rate and round-trip delay,especially in difficult mode,which provides an effective solution for low-delay streaming media transmission.

关 键 词：WebRTC 低延迟 流媒体 

分 类 号：TP39[自动化与计算机技术—计算机应用技术]