Some days ago I was reading 3G-HF transmissions on 11430.0 KHz/USB and one of them caught my attention because of its duration, more than 16 seconds (~16208 msec): too long to be an HDL (max 10240 msec for HDL24) or an LDL (max 7198 msec for LDL) transfer.
The Cross Correlation Function in Fig. 1 reveals it consists of 16 frames sent contiguously, each frame has a duration of 1013 msec and consists of 2432 (PSK-8) symbols at 2400Bd speed: although its total duration is misleading, its characteristics match the STANAG-4538 BW-5 burst but transferred 16 times in a row. (Fig. 2).
 |
| Fig. 1 - CCF function |
 |
| Fig. 2 |
Burst Waveform 5 is used by the fast link setup (FLSU) protocol and its closely associated FTM protocol, which are sometimes referred to together as simply FLSU. Their PDUs are distinguished by the protocol field in the first three bits of the 50-bit paylod: "001" for an FLSU PDU versus "100" for an FTM PDU. After decoding, the expected sixteen 50-bit PDUs are shown in Fig. 3. Curiously, only the last contains a valid 3-bit identifier in the protocol field ("001" = FLSU) while the first fifteen PDUs contain the "000" identifier which makes the carried payload at least unusual if not unknown, at least according to the STANAG-4538 first edition.
![]()
The signal has a good quality, no noise or fadings and then its main features are easily obtainable (see the constellation below) so a decoder error appears at least unlikely, even more in the same positions. Moreover, the first 15 PDUs exhibit a sort of altenate patterns which are best visible in Fig. 4 and likely make sense.
 |
| Fig. 4 |
Since the manufacturers of equipments or protocol developers/implementers may modify the standard if it does not impair its performance, perhaps we are dealing with a modification or an extensionof the (Async ?) FLSU PDU, since BW-5 waveform is unchanged. But it's only a my opinion, so comments and hints are welcome!
Since the good quality, I also demodulated the PSK-8 constellation in order to examine the on-the-air symbols and their arrangement. The obtained bitstream has a 768-bit period that makes 256 PSK-8 symbols (Fig. 6). The lenght of the period (256 symbols) is in contrast with the result of 2432 symbols seen above: likely the CCF spikes are synced with the 240 mesc preamble while the lenght of the period is due to the length of the interleaving or the 16-ary ortogonal Walsh function used in BW-5 waveform for the data format.
![]()
https://yadi.sk/d/BhP9ocMt36fCVd
Since the good quality, I also demodulated the PSK-8 constellation in order to examine the on-the-air symbols and their arrangement. The obtained bitstream has a 768-bit period that makes 256 PSK-8 symbols (Fig. 6). The lenght of the period (256 symbols) is in contrast with the result of 2432 symbols seen above: likely the CCF spikes are synced with the 240 mesc preamble while the lenght of the period is due to the length of the interleaving or the 16-ary ortogonal Walsh function used in BW-5 waveform for the data format.
 |
| Fig. 5 |
 |
| Fig. 6 |
https://yadi.sk/d/BhP9ocMt36fCVd