Re: [webrtc-extensions] Network Tokens and WebRTC QoS (#68)

Hi Justin,

We can pass the network token from the application to WebRTC in a similar way that ICE configuration is passed down - through an RTCConfiguration object.  We add an extra field to the configuration object within the WebRTC library, then add a networkTokens parameter when we create a new RTCPeerConnection in the application, and from there trickle it down to STUN/TURN. Eventually we insert tokens at STUN packets and verify them at the network. 

We have an initial implementation using Jitsi and WebRTC, and you can see a short demo here ( at 16:57 --> https://youtu.be/hmiDhiMyGeQ?t=1017 ) . The token  itself is user-centric and application-agnostic, and the actual access control takes place by defining how one can authenticate and fetch a token. It’s trivial to implement different options for this, e.g.,  user-centric for approved applications, user-centric and application agnostic, or application specific by using standard authentication mechanisms to fetch tokens. 

Not an expert on WebRTC by no means, so appreciate any feedback whether there a better way to do this. 

-- 
GitHub Notification of comment by yiannisy
Please view or discuss this issue at https://github.com/w3c/webrtc-extensions/issues/68#issuecomment-772128119 using your GitHub account


-- 
Sent via github-notify-ml as configured in https://github.com/w3c/github-notify-ml-config

Received on Wednesday, 3 February 2021 01:04:45 UTC