3 ms·

There's no voice quality difference, if the packet goes through, that's true. The problem is that without encryption, P25 packets can handle a bit of radio noi

There's no voice quality difference, if the packet goes through, that's true. The problem is that without encryption, P25 packets can handle a bit of radio noise and still be error corrected into usability. With encryption, in practice, certain bits get altered that cannot be error corrected and the packet cannot be decoded. P25 error correction ability is not evenly across the entire packet. HTTP and HTTPS are riding on top of TCP, which mostly guarantees in order bytes, the correct bytes, and no missing bytes. If you loose the guarantees of TCP, and allow say 1 out of 2000 bytes to be altered randomly, HTTP would still mostly work, and HTTPS would be dead as a doornail.

blantonl 9y ago
Again, that is incorrect. Error correction is applied after encryption via 2.8kbs /sec of forward error correction paired with 2.4kbs /sec of data signaling. And the IMBE vocoder also provides it's own error correction. The signaling, voice and error correction is then formatted into P25 speech frames. A transmission starts with a header data unit, identifying the destination of the call and if encryption was used or not. It then continues with a series of logical link data units (LDUs), that carry the digital speech. The LDUs also repeat the signaling information and the synchronization sequence, so radios can join the call late, or reenter if they drop out due to bad RF coverage and then come back into the call. The call ends with a terminator data unit, indicating that the PTT’s been released, and this is the end of the call. To get around this, sometimes just the dispatchers will have encryption on, since those packets are traveling over fiber to the transmission towers, where they are broadcast at high power levels. The other direction, personal radio to repeater tower, is almost never encrypted because of the much lower power of the personal radios. This is also not correct. Encryption can either be manually turned on or off by the individual user (radio) and/or the dispatcher, or encryption can be "strapped" to a talkgroup in cases where full time encryption is used. THe scenario you describe is all based on the user's choice, not an operational need due to power levels.
- danielvf 9y ago
  Here's a paper with the specific details of why P25 has troubles with reliability while encrypted: http://www.crypto.com/papers/p25sec.pdf http://www.crypto.com/papers/p25sec.pdf "Because of the above-described property of the error correction mechanisms used, especially in voice frames such as the LDU1 and LDU2 frame types, there is no mechanism to detect errors in certain portions of transmitted frames. This was a deliberate design choice, to permit undetected corruption of portions of the frame that are less important for intelligibility."

blantonl 9y ago

Again, that is incorrect. Error correction is applied after encryption via 2.8kbs /sec of forward error correction paired with 2.4kbs /sec of data signaling. And the IMBE vocoder also provides it's own error correction. The signaling, voice and error correction is then formatted into P25 speech frames. A transmission starts with a header data unit, identifying the destination of the call and if encryption was used or not. It then continues with a series of logical link data units (LDUs), that carry the digital speech. The LDUs also repeat the signaling information and the synchronization sequence, so radios can join the call late, or reenter if they drop out due to bad RF coverage and then come back into the call. The call ends with a terminator data unit, indicating that the PTT’s been released, and this is the end of the call. To get around this, sometimes just the dispatchers will have encryption on, since those packets are traveling over fiber to the transmission towers, where they are broadcast at high power levels. The other direction, personal radio to repeater tower, is almost never encrypted because of the much lower power of the personal radios. This is also not correct. Encryption can either be manually turned on or off by the individual user (radio) and/or the dispatcher, or encryption can be "strapped" to a talkgroup in cases where full time encryption is used. THe scenario you describe is all based on the user's choice, not an operational need due to power levels.

danielvf 9y ago
Here's a paper with the specific details of why P25 has troubles with reliability while encrypted: http://www.crypto.com/papers/p25sec.pdf http://www.crypto.com/papers/p25sec.pdf "Because of the above-described property of the error correction mechanisms used, especially in voice frames such as the LDU1 and LDU2 frame types, there is no mechanism to detect errors in certain portions of transmitted frames. This was a deliberate design choice, to permit undetected corruption of portions of the frame that are less important for intelligibility."