RE: on the point of mobility & multihoming

["Kanchei Loa" <loa@ieee.org>]

> > Multiaddress mulithoming and mobility solutions must incorporate the
> > following mechanisms:
> >
> > 1. multiplexing in the presence of more than one address pair
> > 2. adding/removing addresses
>
> Well, there is a difference in the requirements in this point in mobility
> and multihoming.
> Both needs a handover mechanism, to stop using a locator and
> start using an
> alternative one, but strictly speaking, the locator set involved in a
> multihoming solution can be fixed, becuase it is known a priori, for
> instance consider noid. In mobility, this is not so, because you
> don't know
> which will be your future locators.
>

That might not be the case in non-mobile dynamic ad hoc connection for exit
routers as specified in Dave Croker's scenario, whose locator set is not
fixed and known priori. The dynamic comes from the nature of the network
topology (wireless), not from mobility. If the line between exit router and
ISP is changed to dotted lines (wireless) instead of one solid line (wired)
in Geoff's "An Architectural View of Multi6 proposals" presentation, then
the difference between mobility and multihoming is almost null.


---------
Kanchei Loa
loa@ieee.org