[Sls-slp] FW: [EXTERNAL] "No limit" on Expedited frame repetitions of USLP?

Tue Oct 22 09:57:30 UTC 2019

Dear SLP WG,

Please take a look at this email and see if it is necessary to change the USLP Blue Book or not.

Thanks!
Greg

From: Wolfgang Hell <wo_._he at t-online.de>
Reply-To: "wo_._he at t-online.de" <wo_._he at t-online.de>
Date: Thursday, July 4, 2019 at 7:18 AM
To: "Kazz, Greg J (312B)" <greg.j.kazz at jpl.nasa.gov>, John Pietras <john.pietras at gst.com>
Cc: "Holger.Dreihahn at esa.int" <Holger.Dreihahn at esa.int>, "Pham, Timothy T (3300)" <timothy.t.pham at jpl.nasa.gov>, "Greenberg, Edward (312B)" <edward.greenberg at jpl.nasa.gov>
Subject: Re: [EXTERNAL] "No limit" on Expedited frame repetitions of USLP?

Hi Greg,

As John, I'm working on the specification of "Functional Resources" which in simple terms are an abstraction of the functions implemented by a ground station and the associated parameters, events and directives. In this context the space link protocols and the associated managed parameters play a major role. Right now John and I are exploring if a common set of Functional Resources can be established that covers the case of fixed length frames as specified in CCSDS 732.0 and CCSDS 732.1 and a different set that covers variable length frames as specified in CCSDS 232.0 and CCSDS 732.1. At least for now our focus is on the forward link and we disregard CCSDS 211.x (Proximity-1).

As you will have understood from John's email below, we are struggling with the way USLP specifies the repetition of frames. Unfortunately your response to John's email has not clarified the issues we are struggling with.

When it comes to repetitions, CCSDS 232.0 is pretty clear in that respect, although I personally would have liked it better not to permit the repetition of CLTUs in case they carry more than one frame. This option is discouraged in Notes, but still permissible. If I read the USLP book correctly, the Maximum Number of Transfer Frames Given to the Coding and Synchronization Sublayer as a Single Data Unit is 1 and therefore what I dislike about 232.0 has vanished in 732.1. (Only on the side: What is the purpose of a managed parameter when this parameter can anyway only have a single value?)
There is a good reason for CCSDS 232.0 not to permit repetition of CLTUs containing one or more BD frames. Looking into the procedure at the receiving end (section 4.4 of CCSDS 232.0) it is obvious that each BD-frame that passes the Frame Acceptance Checks specified in section 4.4.8.3.3 will be accepted. If BD frames were repeated then the telecommands contained would be carried out as many times as such BD frame is successfully received by the spacecraft. This is NOT the desired effect and therefore repetition of BD frames is not supported. BC frames (as BD frames) are subject to the basic Frame Acceptance Checks only, i.e., repetition of BC frames may result in duplicate protocol control commands. However, as opposed to BD frames, where such duplicates may be harmful to the spacecraft, setting the FARM parameters several times to the same value won't do any harm. For AD frames, the FARM at the receiving end will discard any duplicate frames and therefore, even if such frame is sent several times, (at most) one of these frames will be passed on and consequently the embedded command will be executed (at most) once. Therefore the repetition of AD (and BC) frames can be used to increase the likelihood of a command being received successfully by the spacecraft when the link budget is poor and COP-1 cannot be used due to the too long two-way light time.

Subsection 4.2.10.8 of USLP states: "When reference [6] (this is 232.1) is used as the Synchronization and Channel Coding Sublayer, the All Frames Generation Function may request the Synchronization and Channel Coding Sublayer to perform systematic retransmissions of a data unit as described in 2.4.2, unless the data unit contains a USLP Frame carrying service data on the Expedited Service." This means that as expected USLP does not permit the repetition of BD frames, however, it also rules out the repetition of BC frames which is different from CCSDS 232.0 which per VC permits the setting of dedicated limits for AD and BC frames, respectively. Is this difference intended or should subsection 4.2.10.8 be modified? I guess the latter, because USLP has a dedicated managed parameter to limit the repetition of BC frames on a given VC and the last but one paragraph in section 2.4.2 states: "For each VC, management sets the value to be used for the Repetitions parameter when requesting the transfer of USLP Frames carrying SDUs on the Sequence-Controlled Service. For each VC, management sets a similar parameter for USLP Frames carrying COP Control Commands."

If my above assumption that in terms of repetition USLP and TC shall be equivalent is correct, then the last paragraph of section 2.4.2 confuses me. It states: "When requesting the transfer of USLP Frames carrying SDUs on the Expedited Service, USLP does not limit the value of the Repetitions parameter." As discussed above, the repetition parameter for BC frames is limited to a VC-specific value of a dedicated managed parameter and the repetition of BD frames is not permitted, i.e., the repetition parameter is limited to 1. It is true that the FARM cannot be driven into the "lockout" state by means of "Bypass" frames because the VC frame count value of Bypass frames is meaningless, at least for the FARM and in that regard limiting the repetition parameter is not necessary. But is that aspect the intended message of the above quoted statement?

Thank you in advance for clarifying the above listed points. Chances are that we'll get back to you with further USLP questions.

Best regards,
Wolfgang

Am 02.07.2019 um 00:50 schrieb Kazz, Greg J (312B):
Hi John,

I talked this over with Ed Greenberg and we both came to the same conclusion.
In USLP, there is no limit on how many times the set of repeated frames may be sent.  As long as the frame numbers do not violate the acceptance criteria; by that we mean that the reception of a frame does not cause the COP to lockup.    The number of repetitions is an operational matter:  a poor quality link might influence the sending of a higher number of repeats, but that reduces throughput.

I agree with you about the addition of the two notes you propose below. I will let both Gian Paolo and Tom Gannett know about those and those can be added as corrigenda when appropriate.

Thanks!
Greg

From: John Pietras <john.pietras at gst.com><mailto:john.pietras at gst.com>
Date: Monday, July 1, 2019 at 12:36 PM
To: Greg Kazz <greg.j.kazz at jpl.nasa.gov><mailto:greg.j.kazz at jpl.nasa.gov>
Cc: Wolfgang Hell <wo_._he at t-online.de><mailto:wo_._he at t-online.de>, "Holger.Dreihahn at esa.int"<mailto:Holger.Dreihahn at esa.int> <Holger.Dreihahn at esa.int><mailto:Holger.Dreihahn at esa.int>, "Pham, Timothy T (3300)" <timothy.t.pham at jpl.nasa.gov><mailto:timothy.t.pham at jpl.nasa.gov>
Subject: [EXTERNAL] "No limit" on Expedited frame repetitions of USLP?

Hi, Greg.
The last paragraph of 2.4.2 (Systematic Retransmissions) of the USLP Blue Book states
“When requesting the transfer of USLP Frames carrying SDUs on the Expedited Service, USLP does not limit the value of the Repetitions parameter.”

I find this statement to be very misleading. My understanding is that Expedited (BD) frames *cannot* be repeated, although this is itself ambiguous in the USLP book and I arrived at this conclusion from (a) the fact that the TC SDLP book says so (“When requesting the transfer of frames carrying service data units on the Type-B Service, the TC Space Data Link Protocol always sets the value of the Repetitions parameter to one” – 2.4.2 of CCSDS 232.0) and (b) there are only Repetition managed parameters for AD and BC frames.

However, if it *is* true that USLP supports repetitions of BD frames, what does it mean that “USLP does not limit” them?

While I’m on the topic of repetitions of USLP frames, there should be an additional NOTE in table 5-1 that states that the ‘Maximum Value of the Repetitions Parameter to the Coding and Synchronization Sublayer’ parameter is applicable only when that Coding and Synchronization Sublayer is reference [6] (TC Sync and Channel Coding. Similarly, there should be an additional NOTE in table 5-3 that states that the ‘Value for the Repetitions parameter to the Coding Sublayer when transferring USLP Frames carrying service data on the Sequence-Controlled Service’ and the ‘Value for the Repetitions parameter to the Coding Sublayer when transferring USLP Frames carrying COP Control Commands’ parameters are applicable only when the Coding Sublayer is reference [6]. I know that this is already a Blue Book but I suggest that it go into an errata sheet to update when you have other changes to make.

Finally, I find it very interesting (and frustrating) that there are no *normative* requirements for either TC SDLP or USLP to actually set the value of the Repetitions parameter – the only mentions are in the *informative* sections 2.4.2 of each book and in the lists of managed parameters.

Thanks.

Best regards,
John

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.ccsds.org/pipermail/sls-slp/attachments/20191022/4869c81b/attachment-0001.html>