RE: [manet] Re: MANET definition of a neighbor

"Maybe a new stack architecture is needed for MANETS where protocols can
still partition layer specific items yet share cross applicable
information."

I agree, but we may still be able to pull it off without having to mess
up the stack.

John Mullen 

John P. Mullen, Ph.D.
(505) 646-2958
jomullen@nmsu.edu

-----Original Message-----
From: DANIEL BYRNE [mailto:daniel.byrne@adtran.com] 
Sent: Sunday, May 07, 2006 12:18 PM
To: Mullen, John; manet@ietf.org
Subject: RE: [manet] Re: MANET definition of a neighbor

"...That is why I feel a cross-layer approach will win out in the end.
The physical and link layers will have information that is normally
hidden from the routing layer that will be needed to evaluate potential
links, but the MANET protocol will have the job of finding an optimal
network solution."

True, the routing protocol could use information attained by lower
layers.  A mechanism to pass information up the stack is needed, or the
protocol needs to be multi-layer aware. However, the later causes
problems when there are changes in technologies.  Maybe a new stack
architecture is needed for MANETS where protocols can still partition
layer specific items yet share cross applicable information.

_______________________________________________
manet mailing list
manet@ietf.org
https://www1.ietf.org/mailman/listinfo/manet