3 ms·
That quoted voip paper isn't actually as damaging as it sounds. IIRC that 0.6 rating was for less than half of the words so if you're trying to listen to a conv
by hueving 10y ago
That quoted voip paper isn't actually as damaging as it sounds. IIRC that 0.6 rating was for less than half of the words so if you're trying to listen to a conversation to get something meaningful, it's probably not going to happen.
- ontoillogical 10y agoThat's a good point, and the .6 is an optimistic score (I think it was from running a subset of their model) I got the sense that the authors felt that this proves an attack of this sort was possible and viable, but that their model wasn't quite there yet. OTOH this should be enough to acknowledge that the voice encryption/compression scheme they are attacking is not secure.
- cvwright 10y agoFair enough. The larger point, however, is that this stuff was supposed to be encrypted. The adversary shouldn't be able to learn even one bit about the plaintext. After all, nobody would buy a new encryption module that advertised it could protect you 40% of the time.