Defeating Skype Encryption Without a Key | Hacker News Reader

Defeating Skype Encryption Without a Key

securityweek.com

30 pointsprivacyguru4 comments

Open articleView on HN

Edited form of my comment from http://news.ycombinator.com/item?id=2590731, which discussed the same finding:

- Skype's encryption algorithm is not, itself, broken;

- Skype uses a "prediction" algorithm (LPC) to compress voice streams;

- therefore, words etc. have a specific pattern of bandwidth use;

- bandwidth use of encrypted compressed data is equal to bandwidth use of data that has only been compressed;

- these patterns (which can be detected "through" the encryption) allow fairly good reconstruction of the voice stream.

If you like this kind of thing, Google "ssh keystroke timing attacks", or, more generally, "traffic analysis".

The article isn't really clear, you need to read the paper. Quick summary:

They can use the size of the packets to find boundaries to different phonemes, which can then be used to check for e.g. what language is used, or even against known phrases. They have also had some success at determining which phonemes are used based solely on the length of the VBR packets.

Interesting. They are using frequency analysis[1] of LPC phonemes rather than characters. This implies the sound is not encrypted at all (security by obscurity). It definitely implies the (voice) data stream is not being encrypted as a stream.

[1] http://en.wikipedia.org/wiki/Frequency_analysis

Streams are encrypted. They are exploiting VBR(variable bit rate) part of the Skype's architecture. Skype uses VBR to make sure they pack enough fidelity while maintaining the low bandwidth requirements.

It is an interesting paper to read. Initially I thought they exploited LPC which is the base for all our mobile communications as well. As long as VOIP provider doesn't opt for VBR compression, their networks are immune from this attack.

From the paper.

Under VBR, the size of the codebook entry, and thus the size of the encoded frame, can vary based on the complexity of the input frame. The speciﬁcation for Secure RTP (SRTP) [3] does not alter the size of the original payload; thus encoded frame sizes are preserved across the cryptographic layer. The size of the encrypted packet therefore reﬂects properties of the input signal; it is exactly this correlation that our approach leverages to model phonemes as sequences of lengths of encrypted packets.

Also, they included the solution.

For this reason, the use of constant bit-rate codecs is important to consider as an alternative to simple block ciphers for VoIP, since such codecs might improve call quality given a relatively large ﬁxed packet size. Another alternative might even be to drop or pad packets [19, 27, 57], though, in that case, the effect on perceived call quality is unclear. We note, however, that VoIP providers have made no move to employ any such measures: Skype’s SILK, for instance, is a VBR codec. Similarly, one of the leading proposals for 4G, the LTE Advanced standard, speciﬁes a VBR codec for audio [1] and the use of SRTP to secure voice data channels.

Reply on news.ycombinator.com