CPDLC PSX VS CPDLC Hoppie

[Jeroen Hoppenbrouwers]

Just to not cause an after-you/after-you problem: I am currently NOT working on CPDLC for PSX, as I understand that there is a likely better option, which is to fork PSX native CPDLC to the Hoppie's ACARS backend. And I have not been looking at this option yet while John and friends have.

Let me know what you recommend, although I cannot promise any work in a definite time frame. At work I am deeply into the rabbit hole of Iridium ACARS improvements which absorbs most brain capacity and I am terrible at parallel projects.


Hoppie

[John Golin]

I hope to have some time to review this at WorldFlight next week... to at least check the logic will work without any additional changes from Hardy...

[Hardy Heinlin]

By the way, I have a new plan (after World Flight). Instead of adding an extra inhibit Qi, I'll just extend this existing selector on the Instructor's CPDLC console:

(o) Manual control
(o) Random control - very restrictive
(o) Random control - less restrictive
(o) Random control - no restrictions
(o) External control

This allows the user to reset the control mode in case the add-on exits without an autoreset. I.e. the inhibit mode will be visible to the user instead of being hidden in the network background. Just like the "Externally controlled" switch on the Traffic page.


|-|

[John Golin]

Good thinking!

[Hardy Heinlin]

You guys said it takes several minutes until the logon is established. In which phase exactly is this delay?

Phase 1: "SENDING" is displayed ... how long?
Phase 2: "sent" is displayed ... how long?


|-|ardy

[Jeroen Hoppenbrouwers]

Given that my real-world system does CPDLC all day long...

To send anything while over deep water using satellites takes at least 5 and usually about 15 seconds. So the "sending" phase is about that long.

Then the message arrives at the ATC center and somebody or something needs to react. If the logon is automated, this is usually a matter of a second. If a controller needs to take action, it can be up to 30 seconds.

Then the ack needs to come back to the aircraft, which again takes between 5 and 15 seconds.

This all assumes that there is no other ACARS traffic in the way. Add 30 seconds if there is another message in the way.

ACARS over SATCOM isn't the speed monster people assume it is...

Hoppie

[Hardy Heinlin]

So the afore-mentioned delay of several MINUTES in VATSIM occurs after "SENT", not before "SENT".

Is the "sent" event the confirmation of the ACK event?

[Jeroen Hoppenbrouwers]

Could be a couple of minutes.  The interaction with Hoppie by clients is at an interval specified by the client - 1 minute recommended - so a delay before the ATC client even gets the request.

Actually you have a bit more freedom there.

If you are sitting idle and poll for new unsollicited messages, I recommend a poll interval around that one minute. But if you just sent a downlink and you expect an answer, you can poll after 10 or 15 seconds without any problem. Just be reasonable and everything will be fine.

It won't speed up the ATC client receiving your message of course but it won't also delay your own client much. No worst case of two minutes or more.

That said, such a delay is not uncommon in the real world.

The benchmark for oceanic CPDLC is RCP240. A full four minutes between the ground sending a new clearance and the pilots answering. This is including all technical latency and the pilot's reaction time. 99.9% of all CPDLC transactions (ground to air and back) need to be completed within these 240 seconds for an airplane/fleet/type to be allowed to file flight plans indicating CPDLC capable.

Domestic CPDLC uses RCP130 -- 130 seconds to get 99.9% of all clearances answered. "This is a challenge."

Hoppie

[Jeroen Hoppenbrouwers]

Is the "sent" event the confirmation of the ACK event?

90% sure it is. There is a tech ack at the ACARS level that indicates the ground received the request. There may also be a CPDLC ack coming later that is an indication from the end system (ATC) that they received the logon request, but that does not contain the actual accept/reject yet.

I will dig into some realworld CPDLC logs from a 747-400 I have at work what exactly happens there, just for information.


Hoppie

[Hardy Heinlin]

External CPDLC control is now available in PSX update 10.158:

https://aerowinx.com/board/index.php/topic,4191.0.html


|-|ardy


Documentation upon request.

[Captain_Crow]

Requesting documentation pls...
Steff

[Gary Oliver]

Requesting documentation pls...
Steff

We have ourselves a volunteer!!

[Captain_Crow]

So, as far as I understand, there is an external software in work ?
Steff

[Hardy Heinlin]

Documentation:

https://aerowinx.com/board/index.php/topic,6944.0.html


|-|ardy