[aadl]: [AADL Forum] Proposal of Transport_Protocol property

[Derrick Lau]

What was the consensus regarding this proposal?

Sent from Outlook<http://aka.ms/Ox5hz3>




On Wed, Nov 18, 2015 at 12:12 PM -0800, "Hall, Brendan (MN65)" <Brendan.Hall at honeywell.com<mailto:Brendan.Hall at honeywell.com>> wrote:

Sorry for my tardy response here. The following was stuck in the outbox of an old lap-top

Although I can see great value in the provisioning for protocol properties, I feel that we need to address this aspect in a more holistic manner.  Do we need to a framework that will allow for the capture of protocol refinement?
Within the networking annex working group we have discussed the issues of end-to-end safety protocols.
These are commonly deployed to augment the integrity of standardized protocol offerings.  Where the network channel has memory, such as switched Ethernet, these protocols may encompass timestamping and additional inline CRC field and associated checking logic.  In aerospace some are proprietary but many of the features are exhibited in adjacent industry offerings
http://open-safety.org/
How are such protocol stacks and the subsequent layered API calls best modeled in AADL?
Should AADL abstract the details? Or should it allow /provision for the protocol details to be model and refined towards the final implementation?
Should such protocols be complimented by and Error Model?
Will the integrated AADL enable for improved consistency checking?
For example can we check the inline integrity mechanisms are sufficient with respect to the system hazards?
I would hope so.  Here is a reference to some good background material in this area http://users.ece.cmu.edu/~koopman/pubs/faa15_tc-14-49.pdf




From: sae-aadl-users-bounces+brendan.hall=honeywell.com at lists.sei.cmu.edu [mailto:sae-aadl-users-bounces+brendan.hall=honeywell.com at lists.sei.cmu.edu] On Behalf Of Julien Delange
Sent: Thursday, November 05, 2015 8:39 AM
To: aadl-forum at lists.sei.cmu.edu; sae-aadl-users at lists.sei.cmu.edu
Subject: [aadl]: Proposal of Transport_Protocol property

Dear AADL users and committee members,

Please find a proposal to extends the deployment properties and include the transport protocol mechanism. It could be an additional characteristic helpful to analyze models or even generate code.
Any comment/suggestion is appreciated.

Julien.

-------------- next part --------------
HTML attachment scrubbed and removed