Re: [dtn-interest] New routing solutions?

[waldir moreira <[e-mail]>]

Thats right, Scott.
I think if we worked the generic routing framework as mentioned by Anders, we would certainly have an idea of the existing core routing algorithms that could be added to ION as to allow the feeding of the contact plan. dLife, SPRINT, SCORP, .. do consider the "ongoing basis" as to have a snapshot of the actual moment regarding relevant contact (considering the different levels of social interaction between nodes). This could probably make the collection of contact inference engines to allow the extended-CGR. The outcome would be a 'universal' routing solution that could be employed in sparse and dense DTN scenarios. 

Waldir 


> From: [e-mail]
> To: [e-mail]
> Date: Thu, 13 Nov 2014 18:58:51 +0000
> Subject: Re: [dtn-interest] New routing solutions?
> 
> Thanks, Stephen, that will be extremely helpful when we get to the point of having something to test.
> 
> The other angle, which I thought Waldir was talking about, is development of software that infers somewhat-probable contacts (from social media interactions, contact history, whatever) on an ongoing basis, that could be integrated with a CGR-based forwarding daemon.  ION provides an open API for inserting contacts into a node's contact plan, so I think this interface can be kept pretty clean.  Notionally, I would be glad to see any number of clever contact inference engines operating concurrently and cooperating in keeping a given node's contact plan as comprehensive and accurate as possible, so that the extended-CGR forwarding based on that probabilistic time-varying topology information can be made as efficient as possible.
> 
> Scott
> ________________________________________
> From: Stephen Farrell [[e-mail]]
> Sent: Thursday, November 13, 2014 9:11 AM
> To: Burleigh, Scott C (312B); [e-mail]
> Subject: Re: [dtn-interest] New routing solutions?
> 
> On 13/11/14 16:57, Burleigh, Scott C (312B) wrote:
> > Yes, for the purposes of the development I was talking about it would
> > be great to have a really capable source of somewhat-probable
> > contacts.
> 
> We put all the 2010 N4C contacts up on a web server. [1]
> Someone (likely Elwyn or me;-) would probably have to help
> deciphering formats but there's a bunch of them there and
> some pictures [2] to help (and the code to generate all
> those from DTN2 logs).
> 
> We also have some others but 2010 is probably the most useful
> set we have for remote area scenarios.
> 
> S.
> 
> [1] http://dtn.dsg.cs.tcd.ie/n4c-summer10/
> [2] http://down.dsg.cs.tcd.ie/summer10/contacts/pairs/pairs.html
> 
> >
> > Scott ________________________________ From: waldir moreira
> > [[e-mail]] Sent: Thursday, November 13, 2014 8:35
> > AM To: Burleigh, Scott C (312B); [e-mail] Subject: RE:
> > [dtn-interest] New routing solutions?
> >
> > Thanks for your reply, Scott.
> >
> > So in other words, you think we should mature more the inference from
> > social-based contacts, as to allow proper application to the CGR
> > contact plan?
> >
> > Waldir
> >
> >
> > ________________________________ From: [e-mail]
> > To: [e-mail] Date: Wed, 12 Nov 2014 20:15:53 +0000
> > Subject: Re: [dtn-interest] New routing solutions?
> >
> > HI, Waldir.  One possibility: we are starting to look at the
> > possibility of basing a generic DTN routing framework on Contact
> > Graph Routing, by extending the CGR notion of a "contact plan" to
> > include not only scheduled contacts but also predicted contacts (that
> > is, enabling association of a probability less than 1.0 with an
> > identified contact).  We think we've got a reasonable concept of how
> > to use somewhat-probable contacts to direct forwarding efficiently in
> > a general way, but that leaves wide open the question of how to
> > populate the contact plan with those somewhat-probable contacts; the
> > problem of inferring such things from social contacts or patterns of
> > movement or whatever remains less-than-comprehensively solved.
> >
> > Scott ________________________________ From: dtn-interest
> > [[e-mail]] on behalf of waldir moreira
> > [[e-mail]] Sent: Wednesday, November 12, 2014 4:16 AM To:
> > [e-mail] Subject: [dtn-interest] New routing solutions?
> >
> > Dear colleagues,
> >
> > Recently our draft on social-aware opportunistic routing has expired
> > (http://www.ietf.org/archive/id/draft-moreira-dlife-04.txt).
> >
> > Despite of a first interest from the DTNRG by the time of its first
> > appearance, our draft has not received any technical feedback as
> > expected (it remained as individual submission and no stream
> > defined).
> >
> > Thus, the intention of this email is to check with my colleagues
> > whether there is any interest from the DTNRG (or its IETF WG
> > counterpart) on new routing solutions.
> >
> > I recall that some emails were exchanged about a generic framework to
> > allow better development of new solutions, but that didnt go far as
> > well.
> >
> > Be it for new routing approaches or for a generic routing framework,
> > we would be more than happy to contribute if there is still any
> > interest from this RG in such direction.
> >
> > Best regards,
> >
> > Waldir Aranha Moreira Junior, Ph.D. Researcher at the Association for
> > Research and Development in Cognition and People Centric Computing -
> > COPELABS Invited Professor at University Lusofona - ULHT/ULP External
> > collaborator of the Research Group on Computer Networks and
> > Multimedia Communications - GERCOM/UFPA LinkedIn
> > profile<http://www.linkedin.com/in/waldirmoreira>
> >
> > _______________________________________________ dtn-interest mailing
> > list [e-mail]
> > https://www.irtf.org/mailman/listinfo/dtn-interest
> >
> >
> >
> > _______________________________________________ dtn-interest mailing
> > list [e-mail]
> > https://www.irtf.org/mailman/listinfo/dtn-interest
> >
> 
> _______________________________________________
> dtn-interest mailing list
> [e-mail]
> https://www.irtf.org/mailman/listinfo/dtn-interest