[brite-users] Unified model

Anukool Lakhina anukool at cs.bu.edu
Wed Aug 13 14:19:12 EDT 2003


Its an open research question on what exactly is a good hierarchical
model. What Alberto was referring to is that once this a good hierarchical
model is known, one can implement it in BRITE easily.

Anukool

On Wed, 13 Aug 2003, Vladimir Blagojevic wrote:

> Alberto,
> 
> 
> > >Can't you use BRITE's top-down hierarchical generation model where AS-level
> > >has been generated using Barabasi-Albert model and router-level by some
> > >other available model and thus achieve desired hiearchical properties
> > >targeted by GT-ITM as well as causality aware degree connectivity
> > >properties achieved by Barabasi-Albert approach?
> > >
> >
> > The models implemented currently in BRITE, as they are, do not
> > necessarily generate topologies matching the properties of GT-ITM TS
> > tooplogies. The approach you described could be used but a new model
> > needs to be created which specifically defines which ASs are transit and
> > which are stubs. BRITEmodels do not do that. Creating such a model
> > should not be difficult building on the extensibility design principles
> > of BRITE
> 
> Does this mean that the unified model as defined at the end of the section
> 2.1 is just a matter of implementation? I thought that developing such a
> unified model is still the topic of research?
> 
> 
> Best,
> Vladimir
> _______________________________________________
> brite-users mailing list
> brite-users at cs.bu.edu
> http://www.cs.bu.edu/mailman/listinfo/brite-users
> 



More information about the brite-users mailing list