[Ns-bugs] [Bug 511] Need ability to change mobility model

[code@nsnam.ece.gatech.edu]

http://www.nsnam.org/bugzilla/show_bug.cgi?id=511





--- Comment #4 from Gustavo J. A. M. Carneiro <gjcarneiro at gmail.com>  2009-04-16 07:18:00 EDT ---
(In reply to comment #3)
> thanks, I see.
> 
> I think that the challenge with this approach is that it somewhat abandons an
> ns-3 invariant which is that, once created, the whole object topology does not
> change. Specifically, this means that the ConfigStore won't be able to capture
> the parameters of the models you change on the fly. Don't you have a per-node
> object which manages the mobility model changes ? If so, can't you just make
> this per-node object be a MobilityModel subclass ?
> 

I have such an object, but it is in Python, and for performance reasons it
cannot directly be a mobility model.

This is not really topology information, IMHO.  The main point of
HierarchicalMobilityModel is that you can have a node traveling inside a
vehicular node, and I am putting this model to good use.  However, this
HierarchicalMobilityModel is not very useful if a node is always attached to
the vehicle and cannot leave the vehicle or switch to another vehicle.  I know
you agree with me because the attributes Parent and Child had already existed
before this patch :-)

My patch is only adding a special value to the Parent attribute.  Parent ==
NULL is currently invalid; with my patch, Parent == NULL means "take the Child
mobility model as absolute coordinates, relative to the world (0,0,0)".

In other words, with the patch, Parent == NULL has the same effect as Parent ==
StaticMobilityModel(0,0,0), but is simpler and more efficient.


-- 
Configure bugmail: http://www.nsnam.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.