Daniel,
Daniel Park wrote:
By reading the draft I become confused by the term mobility, since there
are several types of mobility. Are you referring to seamless mobility,
the one that let us move one node from one network to another and still
have a continuous session? Or the draft just considers nomadism, where
every time a node moves it needs to establish a new session?
I think you could clarify this aspect in section 2 of the draft.
Mobility in this document does not have any restricted meaning. As
long as we work for 6lowpan mobility, we'd like to glance at all of
possibility described above. Further texts need to be added in the
next version following your input of course.
Another consideration that I would like to see clarified is the one
regarding the concept of Home Network (HN). In normal IPv6, HN takes an
important role in mobility since it's the Home Agent that is responsible
to keep track of the CoA of our MN. In the draft you named it as
original IPv6 gateway (OIG). My issue is who defines the OIG? Since the
lowpan nodes can be randomly placed in scenario, in my opinion there can
not be a lowpan HA concept, due to the mobility and fault tolerance
characterizes of such network.
802.15.4 has a PAN Coordinator to manage to its child nodes. It can be
used for HN. Or as pointed out at the previous reply, it depends on
how to design 6lowpan architecture especially hierarchical routing
architecture.
I agree with your approach, since its the normal IPv6 behavior, my only
concern is regarding the energy requirements of the PAN Coordinator in
order to manage mobility of all its devices. Moreover this means that if
the PAN Coordinator for some reason dies, the MN can not be reached?
So it would be interesting to insert in section 4 one example, where a
mobile node, not located on its primary network, needs to contact a
specific lowpan node, using its Home Address (IPv6).
A bit confused. Are you indicating that mobile node located in away
needs to talk with any lowpan node using its Home Address ? Why not
swapping its new address allocated by a new location for the old Home
Address ? Am I missing anything ?
I also agree, a bit confused... moreover my example can be considered in
the " Mobility across two different 6lowPan networks", so you can ignore
my comment.
I also agree that it's important to refer the mobility role of
reduced-function devices as suggested by previous replies.
Ok,
Daniel Park
Tiago Camilo
_______________________________________________
6lowpan mailing list
[email protected]
https://www1.ietf.org/mailman/listinfo/6lowpan
