Masthead VHF antennas
- Thread starter Hoolie
- Start date
JumbleDuck
Well-Known Member
Of course. That's what error correction does. But, as I said, it does not make a wild guess at the missing stuff. It either reconstructs it with complete accuracy from redundant information included for this purpose or, if too much damage has been done, it doesn't try.
Out at sea, far from land, maybe. As anyone who has used VHF or AIS near land knows, they most certainly are not line-of-sight in practice. Signals reflect, refract and diffract all over the place. Scientific fact.
JumbleDuck
Well-Known Member
Good point. I would expect AIS to use whatever transmission protocol is used for other data over VHF, like DSC. Perhaps I'm being too optimistic.
lw395
Well-Known Member
And wrong.
RichardS
N/A
Nope. I didn't say that .... I said "In fact, the signal is identical as the digital error correction reconstructs the missing data exactly."
There is no approximation or speculation involved as I emphasised in my second post.
My summary of this very interesting thread is that AIS VHF transmissions are fundamentally different to voice VHF transmissions because of the digitial technology used. The anomalies which occur with voice transmissions can result in longer than line-of sight reception but it is very hit and miss so, if it happens, it's more down to good luck and a following wind.
With AIS/carrier free/digital tansmissions, the same anomalies can be leveraged much more effectively, and some form of digital error correction, complex or rudimentary, has been implemented which capitalises on this improved anomaly propagation and can restore data which, with analogue voice transmission, is lost forever and can only be restored by repeating all the data again with the same potential losses.
I assume that the people who developed AIS were well aware of the potential benefits of this transmission system but would also have been aware that although they have achieved much longer and very reliable non-line of sight transmission, ultimately this is a non-critical benefit since in 99% of cases since the danger does not actually crystallise until two vessels are within a line-of-sight distance, even if thick fog means that there is no actual sighting.
This AIS "peculiarity" compared to voice VHF has been referred to many times on these forums but, in the 7 years I have been contributing, this is the first time that I have ever seen a cogent explanation of this phenomenon.
Richard
PaulRainbow
Well-Known Member
Care to explain ?
MikeCC
Well-Known Member
I was referring to the switching requirements of voice VHF compared to a digital device like AIS transceiver. Nothing to do with the nature of the eventual output.
The power buildup time on VHF PTT is much longer than for AIS transmission - milliseconds compared to microseconds. The splitter therefore has to be able to respond more quickly than the type for AIS receive-only. Or at least that's how a manufacturer of these devices explained it to me. How the development/component cost compares to justify the price difference, I have no idea.
lw395
Well-Known Member
Do they not do some switching to achieve sufficient isolation between VHF tx and AIS Rx?
JumbleDuck
Well-Known Member
Ah. Yes. Bugger. Hadn't though of that. You're quite right. I shall now delete my previous post and look away, whistling innocently.
BrianH
Well-Known Member
I have, at last, found the AIS transceiver specification, it is an ITU document: ITU-R M.1371-1, Annex 2.
In it I note that the communication standard covers four layers from the open system interconnection (OSI) model, viz.: layers 1 to 4 (physical layer, link layer, network layer, transport layer). The physical layer is responsible for the transfer of a bit-stream from an originator, out on to the data link and this has a parameter reserved for Forward Error Correction (FEC) but later defines: "Forward error correction is not used."
So we must move on to the data packet itself, which is contained at the link layer and equates to one TDMA slot. The total length of the default packet is 256 bits, of which the AIS data portion is 168 bits. Following those data bits is the Frame Check Sequence (FCS) block of 16 bits. From the specification: "The FCS uses the cyclic redundancy check (CRC) 16-bit polynomial to calculate the checksum as defined in ISO/IEC 3309: 1993. Only the data portion should be included in the CRC calculation".
So we now know the AIS transmission contains a CRC, what does the receiver do with it if its own algorithm result does not match? The specification has a clear answer under heading 3.2.3 Error detection and control:
"Error detection and control should be handled using the CRC polynomial as described in §*3.2.2.6. CRC errors should result in no further action by the AIS."
Just as I postulated earlier in my quoted text, the message is discarded. That means, no error correction on individual AIS messages. It is patently obvious there cannot be with no extra support data within the packet or any mechanism in any other communications layer to do so if there is no FEC support as allocated for but not ultimately implemented - probably due to overhead in high traffic scenarios.
.
Last edited:
