Been toying with this for a while...
ISPs are tracking data using DPI techniques and also by limiting known protocol transmission (probably by issuing RST into the stream)
What if a P2P application were written using http/https as its transport mechanism (sending byte streams in xml for example?)
a) encrypted avoiding DPIb) using http which makes it very difficult to filter from normal web traffic
Would need centrallised server to act as fallback for when low numbers of users connected (redundancy) with swarm interconnecting during high takeup.
? thoughts
I agree
No opinion
I disagree
Add your comment
They aren't limiting by protocol anymore, as I understand. They are filtering by traffic patterns that seem to possibly be bandwidth hogging traffic.
as a UK BT customer I think that there's definitely a DPI algorithm in there.
I say this because the instant I start downloading it caps me if in peak hours. + even with cap on p2p I can download at full speed on http for hours on end.
This says to me that they aren't (BT) checking on usage so much as protocol/ports.