Results 1 to 5 of 5

Thread: Thales T25 audio corruption when receiving encrypted GPS enabled APX signal

  1. #1
    Join Date
    Jun 26, 2019
    Location
    Australia
    Posts
    28
    Thanks
    14
    Thanked 27 Times in 7 Posts
    Country: Australia

    Default Thales T25 audio corruption when receiving encrypted GPS enabled APX signal

    When receiving encrypted audio from an APX with GPS enabled on a Thales 25, the recovered audio is corrupted.

    Turning off GPS on the APX allows for clear encrypted audio reception on the T25. Disabling crypto on the APX whilst still sending GPS also works. (Of course, T25 does not decode or display GPS).

    Ideally, the T25 should ignore GPS, so I don't have to manually disable the feature on the APX via menu when communicating to a T25.

    As the T25 appears to work fine with encryption off, the problem appears to be either the T25's failure to de-mux the encrypted datastream correctly into audio and GPS data, or the APX may be flagging the GPS data stream in a way that is confusing the T25.

    The T25 is running firmware 8.3.1 and the APX is on firmware R15.01.00

    I suspect a T25 firmware bug, but would like to hear from anyone who has experienced this issue, particularly if you have solved it.

    Regards,
    Phil


  2. #2
    Join Date
    Oct 09, 2012
    Posts
    320
    Thanks
    73
    Thanked 289 Times in 121 Posts

    Default

    seeing as how the APX sends the GPS is outside of the standard and did not exist when the T-25 was being made, I do not know that I would call this a T-25 bug. There may be 2 issues.
    1. a change in the standard that happened after the T-25 or
    2. an implementation that does not 100% meet the standard on the APX.

  3. The Following User Says Thank You to radioinstl For This Useful Post:

    Phil (Aug 27, 2019)

  4. #3
    Join Date
    Aug 15, 2019
    Posts
    42
    Thanks
    34
    Thanked 77 Times in 36 Posts
    Country: Australia

    Default

    This 2013 patent seems to describe what Motorola is doing with the APX. (Props to MattSR, who pointed this out to me in an unrelated conversation.)

    Reading the patent, I'm not surprised it breaks things in other vendors' implementations.

    https://patents.google.com/patent/US9232376B2/en

  5. The Following User Says Thank You to syntrx For This Useful Post:

    Phil (Aug 27, 2019)

  6. #4
    Join Date
    Jun 26, 2019
    Location
    Australia
    Posts
    28
    Thanks
    14
    Thanked 27 Times in 7 Posts
    Country: Australia

    Default

    Thanks to you and MattSR for the patent reference.

    Based on my hasty interpretation of the patent, Motorola appear to have tried to design this mechanism to be backwards compatible with a non-Motorola radio that doesn't understand the embedded metadata.

    I tend to agree with radioinstl's pointing to one of two issues. However, I'm still scratching my head as to why Motorola's voice LDU bit stealing doesn't affect a T-25 receiving unencrypted P25 traffic. In theory, the encrypt/decrypt 'should be' decoupled from the upper layer LDU protocol. We should see no difference in behaviour, regardless of the encryption settings.

    I have a theory that the LDU stealing algorithm may be creating LDUs that bypass the encryption engine. The receiving T-25 decrypts them before forwarding them to the vocoder, thereby creating garbled audio. I'm setting up an op25 wireshark tap to test my theory.

    In the meantime, I would be interested to hear from anyone else who has an APX with the GPS feature enabled talking successfully to any other brand radio using both encryption and cleartext.

    btw, the idea of stealing the least significant bit(s) of a lossy digital channel (video/voice) to embed metadata or hidden messages (steganography) has been around for a while.

    I'm pretty damn sure that the LSB stealing idea pre-dates Motorola's patent filing date of 9 Aug 2013 by decades, thereby raising the question as to why the patent was granted in the first place. ;-)

    Regards,
    Phil

  7. #5
    Join Date
    Apr 09, 2012
    Location
    Australia
    Posts
    908
    Thanks
    251
    Thanked 628 Times in 238 Posts
    Country: Australia

    Default

    And props to duggerd who pointed me in the direction of that patent while I was digging into location services!


    Quote Originally Posted by syntrx View Post
    This 2013 patent seems to describe what Motorola is doing with the APX. (Props to MattSR, who pointed this out to me in an unrelated conversation.)

    Reading the patent, I'm not surprised it breaks things in other vendors' implementations.

    https://patents.google.com/patent/US9232376B2/en

  8. The Following User Says Thank You to MattSR For This Useful Post:

    duggerd (Aug 31, 2019)