There is a lot of good work here.  The solutions document provides an
excellent survey of the existing work.  But the analysis seems lacking
especially in areas where the existing work is kind of lacking, such
as security.  I don't believe that this ballot is sufficient as a
basis for a standards-track nemo ro approach.  There are a lot of
discussions of security, architecture and how well considered the
input proposals are that still need to happen.  I think there needs to
be significannt input from the IPV6 and MIP6 communities before we get
to an approach that is likely to withstand IETF review.


I'm balloting an abstain rather than a no-ob because it seems like
there is a lot of half-baked analysis mixed in with some really good
stuff and I'm concerned that separating it out will be hard.  Specific examples that really triggered this abstain include:

The claim in 2.5 of the problem statement that there is a security
problem associated with accepting traffic from visiting nodes.


The idea of using  an anycast address to find correspondant routers.  

The idea of creating a multilink subnet for all the nodes in the
mobile network.

Use of CGA to prove ownership of a prefix.

Yes, these ideas have been brought forward.  They should be discussed.
But the analysis seems rather incomplete.