7 ms·

Ask HN: Good Resources on Voice Encryption?

38 points by quotz 6y ago

Any textbooks, very technical stuff is recommended. Both, VoIP and Cellular networks.

giantg2 6y ago
I don't have anything specific. You could look for resources that cover digital HAM radio operation. They should have some stuff about the basics of voice encryption. Most of it is not secure until you get to high-end stuff like Motorola AES 256. Some of this 'encryption' is just privacy codes (cell networks are not encrypted but use digital privacy codes I think). Once you digitize the voice, then it should be pretty much regular encryption.
- kawfey 6y ago
  This is the opposite of what they should do. Encryption is verboten on amateur radio.
  - giantg2 6y ago
    Like I said, most of it is not true encryption but privacy codes or digital talk group settings. Encryption does happen on amatuer radio even though it's supposed to be illegal. I'm not recommending they use encryption in this fashion, but that there are some resources out there which describes how it works. https://www.amateurradio.com/encryption-is-already-legal-its-the-intention-thats-not/ https://www.amateurradio.com/encryption-is-already-legal-its...
    - jrockway 6y ago
      The author of this article does not know what encryption is. He's mad that proprietary audio codecs can be used on amateur radio bands, but the codecs are documented in the patent applications (since expired) so no encryption is occurring. It's like not knowing that you need an MP3 player to listen to MP3s. That's not encryption, that's not having an MP3 player.
      - giantg2 6y ago
        How would you define encryption? I would say it's encoding of information so that unauthorized people are less likely to have access to the information . Like I said in my original post, you can look into encryption in the amatuer bands. This article is one example. If you've taken encryption courses then you know that XOR can be a cipher, it's just not secure. The same way a lock on your shed is only going to keep curious people out and not real criminals, the same can be said for lesser forms of encryption. Please look into Kenwood AES/DES models that operate in the VHF/UHF amatuer bands if you still have any doubts about the use of encryption. There are various 40, 56, etc bit encryption schemes as well.
        jrockway 6y ago
        Encrypted communication is communication that is indecipherable if you know the encryption algorithm but don't have the key material. Not knowing the algorithm is a separate thing. In the case of D-STAR, the algorithm is well known, it's just not analog FM. People with FM receivers would be just as upset to hear two people talking with SSB or CW, both perfectly legal on VHF. Over on HF, digital modes like FT8 are quite popular. You will never be able to decode that without a computer, but that doesn't make it illegal (or encrypted). The algorithm is well defined and it's ~100x more bandwidth efficient than SSB voice. Kenwood sells AES radios for non-amateur communications. It's perfectly legal for non-amateur licensees to encrypt their radio links. Your cell phone is doing it right now! Yes, you can use those on the amateur bands if you want. It's against the rules. The reason amateur radio exists is because we follow the rules. It can all disappear overnight if you violate them. There are a few well-known bad apples that make it harder for all of us. It's not to be encouraged.
        giantg2 6y ago
        Standard cell phones do not use encryption. You can intercept calls if you have receivers on the right frequency and using the right tech (believe they are trunked). I'm not advocating that they use true encryption in the ham bands, but it does exist. As I said before, there is information out there about how voice is digitized, how psuedo or weak encryption works (DMR, 40 bit, 56 bit), and then you can use standard encryption practices to encrypt it. The OPs question was about where he can find technical info on encrypting voice communications and I said this would be a good starting point. This is a valid contribution to his question. I am sorry if you do not believe that weak encryption (where you know all possible keys, or the keys are small) is not encryption.
aosaigh 6y ago
What exactly are the challenges of voice encryption over say text encryption? Just the bandwidth of data, or establishing trust?
- cnorthwood 6y ago
  Managing latency is a problem, as you can only encrypt fairly short packets at a time and send each one out, rather than the whole message, and the structure of the data being different can lead to information leakage
  - neolog 6y ago
    Interesting. I haven't tried, but I'd think encrypting and decrypting a packet would take less than 0.1ms, no? What's the variation in the structure of the data? Wouldn't you just encrypt and send a fixed-length interval of audio each period?
- corty 6y ago
  Voice compression interacts unfavourably with general-purpose encryption because one can still infer periods of silence and distinguish low-complexity sounds from high-compexity ones. A rolling "r" compresses badly, a long featureless vowel compresses well. That way you can infer all kinds of information about whats being said, who is saying it, language, reaction, pauses, etc. So for voice encryption, you need to obscure all that through artificial jitter and noise, and lack of compression in strategic places. It is a complex topic and I'm not sure the science is settled beyond "skipping compression helps".
  - t-writescode 6y ago
    With good encryption, a file full of zeros should be indistinguishable from a file full of data. There is something to be said about the volume of data being sent during silence; but, uncompressed audio shouldn't have that problem; and, for audio that does, there could be filler data to maintain a given bitrate across the line.
femto 6y ago
Look at some of the narrow band radio standards. P25 standards are not freely available, but DMR standards can be freely downloaded from ETSI [1]. P25 with 56-bit DES is well and truly broken, but I gather 256-bit AES is still okay. Also look at the signal protocol? [1] https://www.etsi.org/technologies/mobile-radio https://www.etsi.org/technologies/mobile-radio
pjc50 6y ago
GSM is this: https://en.wikipedia.org/wiki/A5/1 https://en.wikipedia.org/wiki/A5/1 (over the AMR codec). The 3GPP docs are generally available if you want to get very, very technical about the mobile network. Generally any combination of a constant-bitrate low-latency voice codec with a cipher in XOR stream mode should work. If the codec does "comfort noise" you should disable that to keep the bitrate constant during silence.
- daneel_w 6y ago
  Is it wise to recommend GSM's encryption today, given that it has been completely and definitively broken?
- viraptor 6y ago
  Do you mean stream mode per packet or per connection? When dealing with voice encoding you need to ensure each packet is self contained and you can drop seconds of traffic without issues. They means the naive application of steam ciphers won't work.
withinboredom 6y ago
I remember being on the old Sprint CDMA network, you could dial ##VPON# and it would encrypt your connection to the tower[1] but was disabled by default. Dunno if that still works (I no longer live in the US, and EVDO isn't a thing anymore). I'm fairly certain Cell networks are not encrypted at all, by default. Or at least it's disabled completely by the towers in Afghanistan. :whistling: [1] https://bestcellular.com/dial-codes/ https://bestcellular.com/dial-codes/
- izacus 6y ago
  > I'm fairly certain Cell networks are not encrypted at all, by default. Or at least it's disabled completely by the towers in Afghanistan. :whistling: This is patently false, while some networks did allow no encryption (A0 on GSM for example), pretty much everything on UMTS/LTE is encrypted at least to the tower.
  - withinboredom 6y ago
    Yeah, not sure how we flew around and listened to cell conversations in realtime. Dunno if we just had the keys, encryption was disabled or the device was operating as a MITM tower.
    - hnanon1 6y ago
      I assume, depending on your aircraft, it was your dirtbox putting out a stronger signal operating as a MITM.
    - zokier 6y ago
      GSM encryption is also pretty broken, e.g. > In 2006 Elad Barkan, Eli Biham and Nathan Keller demonstrated attacks against A5/1, A5/3, or even GPRS that allow attackers to tap GSM mobile phone conversations and decrypt them either in real-time, or at any later time https://en.wikipedia.org/wiki/A5/1 https://en.wikipedia.org/wiki/A5/1
    - bladegash 6y ago
      Not to be a jerk, but this seems like a really inappropriate venue to discuss operational details like this. Especially capabilities/vulnerabilities that, if they do exist, would be classified and concerning a location where the US still conducts military operations.
cfstras 6y ago
The mumble protocol is a nicely documented example for using encrypted Opus (OCB-AES128) over UDP: http://mumble-protocol.readthedocs.io/en/latest/voice_data.html http://mumble-protocol.readthedocs.io/en/latest/voice_data.h...
tenebrisalietum 6y ago
If you want to start at the very beginning, look into SIGSALY. https://en.wikipedia.org/wiki/SIGSALY https://en.wikipedia.org/wiki/SIGSALY
daneel_w 6y ago
I would suggest a basic how-to on setting up SRTP/SIPS for e.g. an Asterisk setup - it's not overly complicated, based on my own experience running a community PBX many years ago.
bhaavan 6y ago
https://www.gstatic.com/duo/papers/duo_e2ee.pdf https://www.gstatic.com/duo/papers/duo_e2ee.pdf is a paper on video encryption by duo. It is pretty good. I assume the audio encryption problem is only a subset of this problem.
dahi1203 6y ago
Are you just searching for free netflix to watch series then these method surely can help you to get it for free <a href="https://modlelo.com/netflix-cookies/">netflix https://modlelo.com/netflix-cookies/">netflix cookies</a>
dahi1203 6y ago
Are you a TV shows, movies, web series lover and want to watch it on netflix for free then just download our latest <a href="https://modlelo.com/netflix-mod-apk/">netflix https://modlelo.com/netflix-mod-apk/">netflix mod apk</a>

38 points by quotz 6y ago

Any textbooks, very technical stuff is recommended. Both, VoIP and Cellular networks.

I don't have anything specific. You could look for resources that cover digital HAM radio operation. They should have some stuff about the basics of voice encryption. Most of it is not secure until you get to high-end stuff like Motorola AES 256. Some of this 'encryption' is just privacy codes (cell networks are not encrypted but use digital privacy codes I think). Once you digitize the voice, then it should be pretty much regular encryption.

kawfey 6y ago
This is the opposite of what they should do. Encryption is verboten on amateur radio.
- giantg2 6y ago
  Like I said, most of it is not true encryption but privacy codes or digital talk group settings. Encryption does happen on amatuer radio even though it's supposed to be illegal. I'm not recommending they use encryption in this fashion, but that there are some resources out there which describes how it works. https://www.amateurradio.com/encryption-is-already-legal-its-the-intention-thats-not/ https://www.amateurradio.com/encryption-is-already-legal-its...
  - jrockway 6y ago
    The author of this article does not know what encryption is. He's mad that proprietary audio codecs can be used on amateur radio bands, but the codecs are documented in the patent applications (since expired) so no encryption is occurring. It's like not knowing that you need an MP3 player to listen to MP3s. That's not encryption, that's not having an MP3 player.
    - giantg2 6y ago
      How would you define encryption? I would say it's encoding of information so that unauthorized people are less likely to have access to the information . Like I said in my original post, you can look into encryption in the amatuer bands. This article is one example. If you've taken encryption courses then you know that XOR can be a cipher, it's just not secure. The same way a lock on your shed is only going to keep curious people out and not real criminals, the same can be said for lesser forms of encryption. Please look into Kenwood AES/DES models that operate in the VHF/UHF amatuer bands if you still have any doubts about the use of encryption. There are various 40, 56, etc bit encryption schemes as well.
      - jrockway 6y ago
        Encrypted communication is communication that is indecipherable if you know the encryption algorithm but don't have the key material. Not knowing the algorithm is a separate thing. In the case of D-STAR, the algorithm is well known, it's just not analog FM. People with FM receivers would be just as upset to hear two people talking with SSB or CW, both perfectly legal on VHF. Over on HF, digital modes like FT8 are quite popular. You will never be able to decode that without a computer, but that doesn't make it illegal (or encrypted). The algorithm is well defined and it's ~100x more bandwidth efficient than SSB voice. Kenwood sells AES radios for non-amateur communications. It's perfectly legal for non-amateur licensees to encrypt their radio links. Your cell phone is doing it right now! Yes, you can use those on the amateur bands if you want. It's against the rules. The reason amateur radio exists is because we follow the rules. It can all disappear overnight if you violate them. There are a few well-known bad apples that make it harder for all of us. It's not to be encouraged.
        giantg2 6y ago
        Standard cell phones do not use encryption. You can intercept calls if you have receivers on the right frequency and using the right tech (believe they are trunked). I'm not advocating that they use true encryption in the ham bands, but it does exist. As I said before, there is information out there about how voice is digitized, how psuedo or weak encryption works (DMR, 40 bit, 56 bit), and then you can use standard encryption practices to encrypt it. The OPs question was about where he can find technical info on encrypting voice communications and I said this would be a good starting point. This is a valid contribution to his question. I am sorry if you do not believe that weak encryption (where you know all possible keys, or the keys are small) is not encryption.

aosaigh 6y ago

What exactly are the challenges of voice encryption over say text encryption? Just the bandwidth of data, or establishing trust?

cnorthwood 6y ago
Managing latency is a problem, as you can only encrypt fairly short packets at a time and send each one out, rather than the whole message, and the structure of the data being different can lead to information leakage
- neolog 6y ago
  Interesting. I haven't tried, but I'd think encrypting and decrypting a packet would take less than 0.1ms, no? What's the variation in the structure of the data? Wouldn't you just encrypt and send a fixed-length interval of audio each period?
corty 6y ago
Voice compression interacts unfavourably with general-purpose encryption because one can still infer periods of silence and distinguish low-complexity sounds from high-compexity ones. A rolling "r" compresses badly, a long featureless vowel compresses well. That way you can infer all kinds of information about whats being said, who is saying it, language, reaction, pauses, etc. So for voice encryption, you need to obscure all that through artificial jitter and noise, and lack of compression in strategic places. It is a complex topic and I'm not sure the science is settled beyond "skipping compression helps".
- t-writescode 6y ago
  With good encryption, a file full of zeros should be indistinguishable from a file full of data. There is something to be said about the volume of data being sent during silence; but, uncompressed audio shouldn't have that problem; and, for audio that does, there could be filler data to maintain a given bitrate across the line.

femto 6y ago

Look at some of the narrow band radio standards. P25 standards are not freely available, but DMR standards can be freely downloaded from ETSI [1]. P25 with 56-bit DES is well and truly broken, but I gather 256-bit AES is still okay. Also look at the signal protocol? [1] https://www.etsi.org/technologies/mobile-radio https://www.etsi.org/technologies/mobile-radio

pjc50 6y ago

GSM is this: https://en.wikipedia.org/wiki/A5/1 https://en.wikipedia.org/wiki/A5/1 (over the AMR codec). The 3GPP docs are generally available if you want to get very, very technical about the mobile network. Generally any combination of a constant-bitrate low-latency voice codec with a cipher in XOR stream mode should work. If the codec does "comfort noise" you should disable that to keep the bitrate constant during silence.

daneel_w 6y ago
Is it wise to recommend GSM's encryption today, given that it has been completely and definitively broken?
viraptor 6y ago
Do you mean stream mode per packet or per connection? When dealing with voice encoding you need to ensure each packet is self contained and you can drop seconds of traffic without issues. They means the naive application of steam ciphers won't work.

withinboredom 6y ago

I remember being on the old Sprint CDMA network, you could dial ##VPON# and it would encrypt your connection to the tower[1] but was disabled by default. Dunno if that still works (I no longer live in the US, and EVDO isn't a thing anymore). I'm fairly certain Cell networks are not encrypted at all, by default. Or at least it's disabled completely by the towers in Afghanistan. :whistling: [1] https://bestcellular.com/dial-codes/ https://bestcellular.com/dial-codes/

izacus 6y ago
> I'm fairly certain Cell networks are not encrypted at all, by default. Or at least it's disabled completely by the towers in Afghanistan. :whistling: This is patently false, while some networks did allow no encryption (A0 on GSM for example), pretty much everything on UMTS/LTE is encrypted at least to the tower.
- withinboredom 6y ago
  Yeah, not sure how we flew around and listened to cell conversations in realtime. Dunno if we just had the keys, encryption was disabled or the device was operating as a MITM tower.
  - hnanon1 6y ago
    I assume, depending on your aircraft, it was your dirtbox putting out a stronger signal operating as a MITM.
  - zokier 6y ago
    GSM encryption is also pretty broken, e.g. > In 2006 Elad Barkan, Eli Biham and Nathan Keller demonstrated attacks against A5/1, A5/3, or even GPRS that allow attackers to tap GSM mobile phone conversations and decrypt them either in real-time, or at any later time https://en.wikipedia.org/wiki/A5/1 https://en.wikipedia.org/wiki/A5/1
  - bladegash 6y ago
    Not to be a jerk, but this seems like a really inappropriate venue to discuss operational details like this. Especially capabilities/vulnerabilities that, if they do exist, would be classified and concerning a location where the US still conducts military operations.

cfstras 6y ago

The mumble protocol is a nicely documented example for using encrypted Opus (OCB-AES128) over UDP: http://mumble-protocol.readthedocs.io/en/latest/voice_data.html http://mumble-protocol.readthedocs.io/en/latest/voice_data.h...

tenebrisalietum 6y ago

If you want to start at the very beginning, look into SIGSALY. https://en.wikipedia.org/wiki/SIGSALY https://en.wikipedia.org/wiki/SIGSALY

daneel_w 6y ago

I would suggest a basic how-to on setting up SRTP/SIPS for e.g. an Asterisk setup - it's not overly complicated, based on my own experience running a community PBX many years ago.

bhaavan 6y ago

https://www.gstatic.com/duo/papers/duo_e2ee.pdf https://www.gstatic.com/duo/papers/duo_e2ee.pdf is a paper on video encryption by duo. It is pretty good. I assume the audio encryption problem is only a subset of this problem.

dahi1203 6y ago

Are you just searching for free netflix to watch series then these method surely can help you to get it for free <a href="https://modlelo.com/netflix-cookies/">netflix https://modlelo.com/netflix-cookies/">netflix cookies</a>

dahi1203 6y ago

Are you a TV shows, movies, web series lover and want to watch it on netflix for free then just download our latest <a href="https://modlelo.com/netflix-mod-apk/">netflix https://modlelo.com/netflix-mod-apk/">netflix mod apk</a>