WiFi Stats.

[karthikbr82]

The quality of Wifi directly affects the performance of WebRTC and WebRTC can use these stats to adapt and produce a better experience for the users.

Can we standardize WiFi metrics?

The metrics can be - RSSI, Signal, Noise etc.

[vr000m]

I suppose this is similar in behaviour to the audio level, where the latest or current value at the time of polling is be reported in the stat.

[alvestrand]

I suspect that defining Web accessible metrics for network APIs belongs to the groups dealing with device interfaces (I think it's still the DAP WG), not to WebRTC. https://www.w3.org/TR/netinfo-usecases/ might be a relevant document (from 2014).

[aboba]

Agree with Harald that WiFi stats are out of scope of the WebRTC-stats document. In particular, if you want to understand WiFi performance, it is necessary to go below the IP visible metrics (e.g. packet loss) into link layer metrics relating to frame loss, link layer retransmission and rate adaptation.

[vr000m]

We added networkType in RTCIceCandidateStats in https://github.com/w3c/webrtc-stats/pull/274, and the wireless signal-related metrics might need to be correlated to the metrics reported from the network (e.g., losses and RTT).

@karthikbr82 could you make an issue on https://github.com/WICG/netinfo/issues and see what they say? link to this issue if needed.

[alvestrand]

Iceboxed while waiting for netinfo.