15 February 2018

A duble [5N2]N1 framing?

In case of an encrypted Async transmission the start bit and the n-stop bits that wrap the text are no more recognizable at the crypto device output, or - at  receive side - just after the removal of the HF waveform overhead (Figure 1).
 
Fig. 1
Well, actually this does not happen in case of the Swedish Army 8-bit text transmissions [1]. Indeed, looking at the bitstreams obtained after the 110A removal (regardless their different formats) we see clear 8N1 framings that wrap encrypted texts (Figure 2): as per above, we should not see those start/stop bits but only blocks of chipertext.

Fig. 2

I tend to think that it depends on the flowchart depicted in Figure 4:

Fig.4
Most likely the message is produced by an Async 5N2 system (ACP-127, other MHS,...), then the message is encrypted. The third block, downstream the encryptor block, processes the whole flow adding the header bytes and the so-called Z-string (see the above link). This is an Async 8-bit application/terminal that uses 1 start bit and 1 stop bit framing (8N1) so that the resulting stream sent to the 100A modem is a 10-bit framed stream: ie, a sort of a "5N2 in its turn framed into 8N1" schema.

One could say that the first three functional blocks could be accomplished with a single physical device (the dashed rectangle in Figure 4). In this regard, it's interesting to know that the Swedish Defence Materiel Administration (FMV) and the Swedish Armed Forces, in a joint project with Tutus Data, developed an off-line file encryption and decryption application for Windows called "Filkrypto PGBI" [3]: its typical usage is the secure transfer of sensitive or classified information. This way, the headers and Z-strings could be added by the encryptor block.

Quoting my friend J-S4538, who helps in this work, in the end, we don't know what data they transmit, if it is ACP-127, tactical data or other. It looks like they use two kind of modems or transmission systems, one uses the {}-protocol and STANAG-5066, the other one an old-school 8N1 terminal. Maybe one is for point-to-point the other one for broadcast. Or it is related to two different groups of stations.

Fig.5