Some providers will give unlimited traffic for things like Facebook, Twitter, LinkedIn and even YouTube in some rare cases on smartphones. I'm wondering if there's any documented way they identify this traffic. Not for malicious purposes but from an ethical standpoint I'm curious as to how they keep the data at bay. On first glance, I would assume an IP filter and payload size checks but I wonder if anyone has any better theories or resources?
Strictly speaking, the phones are not completely restricted because of IP over Facebook. Whatever the medium, there's always someone to think about transporting raw IP packets in it, and then go ahead and actually implement it.
For the actual restriction, this can be done with IP filtering, or by enforcing an unescapable Web proxy which filters on the URL (preferably a transparent proxy rather than something in the phone software, so that jailbroken phones may not evade the restriction).