AIS Class B delays

[Roberto]

On my boat I have a combined VHF/AIS receiver, plus an AIS transponder.
The two are totally separate: two gps, two antennas, etc
The radio AIS RX flow is fed into a PC with Opencpn;, the transceiver works by itself, no cables to anything. The signal sent into the air by the transceiver is received by the VHF/AIS antenna, and fed into the PC.

This allows to see what kind of difference exists between my own AIS signal and my actual position.

After a number of days of close observation, I noticed that usually the 30s update time Class B use is more than enough for anti-collision purposes: other boats "see" myself through AIS in a position very near to the actual one.

The problem seems to come out when there is a lot of Class A traffic: in this case, I found differences can be of the order of 0.2-0.3 miles. By "differences" I mean the last position sent by my class B transponder and my actual position. In other words: when my AIS sends my position and speed vector they are of course identical to my actual one, but then they remain for minutes, often having no resemblance with what the boat did later.

Example 1, taken near Finisterre TSS, OpenCpn showed more than one page of AIS targets.
My actual position (red roundish icon), and the last available AIS position my TXR sent (green triangle); this means if others look at their AIS screen they see myself being and doing what the green triangle is doing, whereas I am in the "red" position with the red vector

Example 2
Approaching this port, the delay was such that I had rounded the breakwater and well inside the port and my last AIS position (possibly with a 2-3-4 minutes delay) was still showing my boat outside. Sorry I could not take the screenshot it was too busy.

Ok, AIS positions report their "age", professional shipping tend not to rely on AIS, etc etc but I am tempted to go "stealth" and not send out any ais message in these circumstances. (I have an RTE by the way).

What are your opinions ?

r

 

[Kerenza]

Hi, interesting. To be clear, is it your observation that your class b transmissions are delayed only when class a reception is heavy? It would be interesting to know if that is a model specific phenomenon or common to all class b transmitters.

 

[Roberto]

Hi, interesting. To be clear, is it your observation that your class b transmissions are delayed only when class a reception is heavy?

Yes, I would think so, at least I can not see any other possible cause.
In areas with normal traffic, there are no delays, or at least no delays of navigational relevance: the plotter shows both my gps position and my ais triangle one above the other.


With a lot of class A transmissions (ships, fishing boats etc) the two icons become increasingly separate, I can see on the screen my red boat (fed by the gps data) moving second after second, while my last sent ais position remains fixed behind. Then whenever my transponder manages to send its position of course the green triangle "jumps" forward to match my actual position.

I can see this happening also with ais signals of other sailing boats, presumably class B too: when I look at where they actually are, and where they are shown by their ais signals, there may be relevant differences, example they have already rounded a cape, or entered a port, and their ais signals are still showing position of a few minutes earlier.

It would be interesting to know if that is a model specific phenomenon or common to all class b transmitters.

Indeed. Fwiw, my Ais is an Advansea, I understand the same machine is sold branded under different commercial names.

 

[bitbaltic]

This allows to see what kind of difference exists between my own AIS signal and my actual position.

After a number of days of close observation, I noticed that usually the 30s update time Class B use is more than enough for anti-collision purposes: other boats "see" myself through AIS in a position very near to the actual one.

The problem seems to come out when there is a lot of Class A traffic: in this case, I found differences can be of the order of 0.2-0.3 miles.

Ok, AIS positions report their "age", professional shipping tend not to rely on AIS, etc etc but I am tempted to go "stealth" and not send out any ais message in these circumstances. (I have an RTE by the way).

What are your opinions ?

r

My guess is this is an effect of the AIS TDMA system, which is what organises the time slots for each vessel to broadcast. There are 4500 broadcast slots available in each minute, and If there is a lot of Class A traffic around such that the available slots are getting full, it is effectively prioritised over Class B. Your receiver is basically struggling to detect a slot in which to broadcast your information.

More info in this panel from Digital Yacht:

 

[Kerenza]

I should add of course that class b transmissions are required to monitor class a before sending and transmit only when there is a spare slot - it's the time scale which is the interesting parameter.

Lifted from the Danish maritime authority -

AIS of classes A and B
AIS information is divided into two classes – classes A and B – depending on the AIS transponder transmitting the AIS information. These classes are of great importance to the capabilities of your AIS. There is a great difference between the two classes, both in terms of extent, complexity and price.

AIS information from a class A transponder will always be prioritized and, thus, be shown to other ships in the area, whereas AIS information from a class B transponder will not be shown until or if there is room on the AIS channel.

AIS of class A
In order to avoid that the ships’ AIS systems all speak at the same time, large ships use an AIS system of class A, which is called SOTDMA (Self-Organized TDMA). An algorithm ensures that the AIS transmitter of a ship first notices how other ships transmit their messages and, subsequently, adjusts its own transmission pattern to that of the others. In case there are more ships fitted with AIS of class A in an area than the capacity of the band width, the system will automatically limit the area of coverage so that the remotest ships in the area are discarded.

AIS of class B
Small vessels fitted with AIS, such as recreational craft, can use a less expensive AIS station of class B, which transmits less frequently. This system is called CSTDMA (Carrier Sense TDMA). A class B station will listen for a couple of milliseconds to hear whether a large ship is transmitting before it transmits its own message. Some rather old AIS stations of class A can see only the position, but not the identity of these class B stations – and large ships can choose not to show AIS stations of class B on their displays if their displays become confused because of too many yachtsmen.

Thus, the class B information will be shown to other ships in the area only if or when there is room on the AIS channel. On an AIS transponder of class B you can see whether the information transmitted is prioritized and thus displayed on other ships.

 

[macd]

Thus, the class B information will be shown to other ships in the area only if or when there is room on the AIS channel. On an AIS transponder of class B you can see whether the information transmitted is prioritized and thus displayed on other ships.
[/I]

So TDMA = Too Damn Many AIS signals? (Useful mnemonic, if nothing else.)
But seriously, that's worth knowing. Thanks.

 

[Ludd]

Scary!
It seems to me that at the very time you require accuracy of yoursystem i.e. in busy shipping areas,it actually increases the danger you are in!
This has convinced me NOT to fit AIS other than a receiver.

 

[Roberto]

From what I have seen, distances measured on the chart have been around 0.1-0.2 miles
The problem is the Class B position sent will remain at that spot on the screen until the TXR finds another slot to update it; when clicking on the AIS target icon, the program indicates "Image xxx seconds old" so the important info is there, maybe AIS icons should change colour when older than x seconds?

Oh, I just had a look at Opencpn, there is an option "Mark targets as lost after x (min)", I now set that to 1 minute, so signals older than 1 minute will show the little black bar over the icon. I will check in the next few days if that makes sense.

Still, a bit worrying to know that ships may have our position and speed vector on their screens partially/totally wrong, I suppose they rely on radar plotting on longer distances so no worries (and also fewer signals and a lot more room for class B signals), but in a close crossing situation, say a TSS, port, commercial harbour or similar, I wonder if it isn't better to switch the thing off.

 

[bitbaltic]

I'd be interested to know if there are (or can be) any differences in the way that manufacturers of class B units implement the CSTDMA which might affect how successful it is: ie are some class B units better at 'seeing' an opportunity to broadcast than others, or is it just a fixed method and class B trtansmitters from any manufacturer will perform pretty much the same. Totally outside my knowledge though, can anyone comment?

 

[rogerthebodger]

Scary!
It seems to me that at the very time you require accuracy of yoursystem i.e. in busy shipping areas,it actually increases the danger you are in!
This has convinced me NOT to fit AIS other than a receiver.

That is exactly my view and I always take the view that if he is much bigger than me monitor very closely until you are positive that there will not be any danger and be prepared to take action as per col regs if you consider that the danger has increased to an unacceptable level.

Thinking that the traffic around you know where you are and will avoid you as you as a sailing boat is standon is IMHO suicidal.

Keep a good watch out all the time .

An AIS receiver will allow that over a greater range an AIS transmission on its own will not.

My VHF also displays received AIS data and allows me to make a DSC call to a vessel directly with is MMSI number displayed on the radio screen with CPA and TCPA.

 

[Gypsy]

On my boat I have a combined VHF/AIS receiver, plus an AIS transponder.

After a number of days of close observation, I noticed that usually the 30s update time Class B use is more than enough for anti-collision purposes: other boats "see" myself through AIS in a position very near to the actual one.

r

You don't say what speed you were sailing but the situation is worse if 2kt or less as Class B xmits only every 3mins at these slower speeds.

 

[pvb]

Yet another reason not to waste money on a Class B transceiver! Putting the money towards an active radar transponder buys more certainty and, arguably, potential safety.

 

[MM5AHO]

I had delay problems with a RX only AIS and a slow processor in the laptop.
With a faster processor and a transceiver this issue has gone away. I'm not sure which of the two changes fixed the issue.

 

[Roberto]

more information

as fog and windless days continue, I keep on gathering data


this class B position (not myself, another pleasure boat) shows a delay of about 4 minutes, at say 6 knots-3m/s, 4 minutes = 240 seconds --> difference in position 240*3 --> next to half a mile

there were about 35 AIS targets, mostly class A.

I *think* I have discovered something: while watching the "position age" on the various class B targets, myself included, I have seen that the "ages" before any update are always multiple of 30s.

IMHO, to be verified, a class B sends its position, then waits the regular 30s (if speed above 2kt), at the 30th second it checks if there is a carrier (airwave carrier), if there are none it sends its update message, if it finds a carrier then it puts itself on standyby for another 30 seconds. My impression is it does not monitor constantly what happens on the frequency, trying to send its data as soon as a slot is available, but rather checks every 30s if there is a free slot to send its position, if not then waits 30 more seconds before trying again.
The one above was udated at 4m:30s

 

[jonjo]

Re: more information

I sailed in close company with a previously unknown class B transmitter last night, we shadowed each other for 20 miles hence I had time to watch their transmission frequency. I did not notice missed slots at the western end of Lyme Bay even though my masthead ariel was picking up the dense regions of the TSS off the Casquets. Decent software is going to maintain a projected AIS contact position using data from the last AIS squawk one hopes.

The social benefit of class B AIS is that when the same vessel arrived in Fowey 12 hours after after our close company sail I was able to strike up a pontoon conversation about our voyage experiences with a complete stranger.

 

[Gypsy]

Re: more information

IMHO, to be verified, a class B sends its position, then waits the regular 30s (if speed above 2kt), at the 30th second it checks if there is a carrier (airwave carrier), if there are none it sends its update message, if it finds a carrier then it puts itself on standyby for another 30 seconds. My impression is it does not monitor constantly what happens on the frequency, trying to send its data as soon as a slot is available, but rather checks every 30s if there is a free slot to send its position, if not then waits 30 more seconds before trying again.
The one above was udated at 4m:30s

I think you are right. I recently installed an em-trak B100 transponder. I think, by the look of the Plastimo/Advansea user manual, em-trak may make the TR-210 as an OEM. They both use the ProAIS2 app for setup, monitoring and diagnostics. Have you used this whilst looking in to this issue? ProAis2 has a diagnostic page which logs error messages including failed Tx and timestamps each message. One of the Tx fail messages is 'Busy' (or similar) which indicates it can't find a Tx slot. It will also log a failure to Tx caused by 'No Fix' on the GPS.
It would be very interesting to see how often your system logs a Tx failure message and why.
My B100 has an SD card to save logs to but it might be possible to save to the PC via the USB, I can't recall. I don't see the SD card feature on the TR-210.

 

[Gypsy]

Re: more information

I just reviewed my notes from testing the B100 and found the following explanation by Em-Trak regarding 'CP Busy' messages:
What do the messages “TX Attempt failed” and “CP Busy” mean in ProAIS2?
We’re often asked by concerned customers if their B100 is faulty as they’re seeing messages like “TX Attempt failed (msg 18)” appear in the messages screen of ProAIS2. Any confusion is usually confounded by the fact that the unit appears to be working normally, so what do these messages mean?
Class B AIS (The system on which the B100 operates) is a ‘carrier sense’ system. It has to find a free slot to transmit in by listening just before transmission to check the slot is empty.
When a transmission is scheduled a block of 10 potential slots is selected. It will attempt to transmit in the first of these slots. If the slot is busy the ‘CP busy’ (meaning ‘candidate position busy’) message will be output. It will then try again in the next slot, and so on for the 10 slot block. It is quite normal to see ‘CP busy’ messages in an area where other AIS vessels are operating. It simply means the first slot selected was busy. 99% of the time the transmission will be successful in one of the other 9 slots.
So long as the indicator remains green transmissions are occurring on schedule. If transmissions can’t be made because the AIS environment is simply too busy then the indicator will turn amber (this is incredibly rare though and extremely unlikely to occur outside of lab conditions)
In summary these messages are normal and can be ignored if the indicator is green. Anything you might see in the ‘messages’ box is just for info and not a fault or alarm. If there is an active alarm or fault, then the ProAIS2 will list these in the “Alarms” window of ProAIS2.
From: https://em-trak.com/support/messages-tx-attempt-failed-cp-busy-mean-proais2

As you can see they say it is extremely rare to miss a Tx as they target 10 slots and expect to get one. However, they say the Amber light should come on if Tx failures occur. I suppose they need a few missing Tx before they light the Amber. Can you monitor the LEDs on your TR-210 whilst sailing or doing these tests?

 

[DJE]

The social benefit of class B AIS is that when the same vessel arrived in Fowey 12 hours after after our close company sail I was able to strike up a pontoon conversation about our voyage experiences with a complete stranger.

You want to be careful talking to strange women like her on pontoons. Nice passage though wasn't it?

 

[jonjo]

You want to be careful talking to strange women like her on pontoons. Nice passage though wasn't it?

You have an expert helmswoman there, she landed on the pontoon perfectly.

My excuse for the erratic light signatures she observed at 3am is that past Salcombe I had too much mainsail up and beyond a narrow broadreach groove I started to broach from a Falmouth course and instead pointed at Hope Cove. It certainly ranks as one of my most exhilarating sails, for 3 hours I averaged between 7.5 and 8.5 knots water speed which is too much for a 35ft hull. Who needs the Mediterranean given English Channel nights like that one!

 

[temptress]

Interesting post. We have recently been experiencing AIS delays on our own set and it has gradually been getting worse.

Last week on passage from Tonga to Fiji we unusually sailed in close proximity to another yacht on the same passage and sailing roughly the same speed. As the wind was light and the passage over 300 miles I spent 3 days playing with the AIS.

Sometimes we went 30 minutes without a position from the other yacht and they eventually stopped seeing us on AIS at all.

We were in sight of each other for 3 days with no other vessels around until some fishing boats and ferrys on the last 20.

Chatting over the VHF we noticed that my VHF transmissions were degrading over time as well.

So as I had time I setup the emergency VHF aerial directly to the AIS and later through the splitter and then again to the VHF and any other combination I could think of. This removed the AIS delays and errors I was seeing.

Getting to the point I found a bad connection to the aerial which I remade. It seems that this bad connection while not killing the VHF did cause the AIS problems. Firstly it made transmission difficult and only sometimes did the AIS transmission get through although using ProAIS2 the log file was always reporting a successful transmission. Secondly we did not always receive the signals from the other vessel either.

Check the physical aerial connections on the receiving end of the set-up as this could also cause the things you are seeing.

Good luck and if all else fails just keep a good lookout and carry on sailing.