Just a couple of comments about the different waveforms that can be seen erstwhile analyzing traffic exchanges performed utilizing Thales equipment.
1. As you see in Figure 1, after the proprietary Systeme-3000 Skymaster ALE, data are transferred utilizing STANAG-4285 FEC and the HDR Single speech waveforms (both proprietary) and even STANAG-4539. That's a bit different since, erstwhile link negotiation is complete, the selected traffic waveform shall stay the same during that link session (apart FEC coding, interleaver and data rate).
 |
| Fig. 1 |
The order in which the different waveforms appear does not seem "formalised", most likely it's due to the adaptive feature of the modem - which so adopts different data rates and waveforms - or it's in any way "announced" during the link negotiation. In this regard, the "dual demodulation state" comes to mind, a characteristic of 3G-ALE indicated in STANAG-4538: ie, the nodes partecipating in packet data kind links do not anticipate a same waveform and are ready to demodulate circumstantial xDL waveforms. Looking at Figures 1,2, it seems that shall nodes be required to simultaneously demodulate at most 4 waveforms within the same logical link, i.e. they look for STANAG-4285 FEC and HDR ST (Thales proprietary) or NATO STANAG-4539, and evidently the ALE waveform signaling the link terminate. I don't know how this happens but I would guess the ALE phase announces what waveforms will be used.
 |
| Fig. 2 |
2. As from STANAG-4539 #4.3.1.1, the synchronisation preamble consists of 2 parts. The first part consists of at least N blocks of 184 8-PSK symbols to be utilized exclusively for radio and modem AGC (also known as TLC section). The value of N is configurable to scope from values of 0 to 7. The second section consists of 287 symbols. The first 184 symbols are intended exclusively for synchronisation and Doppler offset removal purposes while the final 103 symbols besides carry information regarding the data rate and interleaver settings. The full duration of the sync preamble as a function of the number N is shown in Table I.
 |
| Table I - sync preamble duration Vs N (at 2400 Bd) |
In all the recordings at my disposal I noticed the same duration of the synchronisation preamble, i.e. 349.5 msec which correspond to N = 3 or 3 184 symbols blocks for the TLC section (Figure 3): this - of course - is just a specified common peculiarity but which at least so far has not been denied in akin recordings (Thales). another S4539 recordings I have analyzed have different preamble sync lengths. As specified above, the value of N is configurable, nevertheless it is not clear if this parameter is accessible to the operator or if it is simply a mill setting (however modifiable).
 |
| Fig. 3 - any Thales STANAG-4539 synchronisation preambles |
Thales engineers surely have the ability to modify the STANAG-4539 preamble, as in the case of the extended preamble employed in their Salamandre HFXL waveforms, nevertheless even military grade modems specified as the Harris RF-5710A are able to admit and demodulate these S4539 bursts, assuming that the data rate and interleaver settings values that are "read" by the modem are actually the exact ones.
 |
| Fig. 4 - my Harris RF-5710A working Thales mixed-mode recordings |
Such further recordings and comments are very welcome.
