4 ms·
I really like this. But how is "typical download speed" determined? It seems extremely non-obvious, particularly since congestion occurs in different spots (e
by crazygringo 4y ago
I really like this.
But how is "typical download speed" determined?
It seems extremely non-obvious, particularly since congestion occurs in different spots (e.g. at the block level or at the apartment building level?) at different times in different locations. In fact I wouldn't even know how to begin to define it.
Is there some kind of methodology for this, either based on real-time monitoring, or else a theoretical calculation based on usage patterns and network topology?
- jjeaff 4y agoFrom what I can gather in one of the linked rule documents, they may not have written the specific rule for determining typical speed yet. It appears to have several comment suggestions and requests additional comment.
- Enginerrrd 4y agoHow about the average speed a user sees normalized by the amount of data for which they see that speed. This biases the result toward evening streaming, but that's typically what people care about anyway.
- crazygringo 4y agoBut how are you going to measure that, is my point. A single user can run a speed test at a given point in time, but you can't be running speed tests all the time. The concept of "the average speed a user sees" -- there isn't an obvious way to me for how an ISP could calculate that for even a single customer, let alone across the whole network. Which is why I'm curious if there are network experts who have ideas of how you could even define it in a way that is both relatively accurate and practical to implement.
- eppp 4y agoOur system can run random speed tests on segments of the fiber onts at various times and produce such reports.
- Enginerrrd 4y ago>But how are you going to measure that, is my point. ISP's seem to have had ZERO issues in the past with metering connections for speed, data, or some combination of those on a per user basis, so it must be possible.
- alkonaut 4y agoI think it seems pretty reasonable. If someone tries to use their broadband on multiple occasions and more often than not doesn't get the advertised typical speed, then it's not typical, and a court should agree with that. I don't really like the idea of making regulation too technical or specific. "Typical" is a good word here.