Article directory
In the previous article WebRTC series-Qos series audio setting packet loss retransmission nack introduced the way to enable audio packet retransmission and the way to verify whether it is turned on successfully; in the article WebRTC series-Qos series RTP/RTCP source code analysis- Section 2.3.3 of RTP/RTCP Packet Analysis-3 introduces how NACK packets are parsed in WebRTC and the definition of the RFC protocol.
In WebRTC, it supports NACK for audio and video. This article only analyzes NACK and its key related RTX mechanism, and does not analyze the relevant knowledge and implementation of jitterbuffer or neteq, which will be introduced in other articles.
1. Brief introduction of NACK
The video in WebRTC is to enable nack by default, and then use RTX to retransmit (the implementation of audio is different, if NACK is enabled, the original ssrc is used to retransmit the package); for RTX, this article will also introduce that RTX will not use the original SSRC to retransmit Instead of passing the lost packet, a new ssrc will be used to retransmit the packet&