Sri,

On Mar 22, 2012, at 12:59 AM, Sri Gundavelli wrote:

> Hi Pierrick,
> 
> Catching up on DMM threads, have been sleeping for a while, some comments
> woke me up :)
> 
> 
> I agree with you. IMO, to achieve DMM we need:
> 
> 1.) Localized Gateway Selection. The ability for the network to assign a
> anchor point closer to the UE location. Ex: Geo-location, RNC ID ... Most
> already got in Rel-10.
> 
> 2.) Once the session is created and the UE moves to a new location. There
> should be a new gateway assignment.  The network should continue to

Assuming most of the applications/communication would be happy with
addresses that come and go within a limited but not too small area..
say a L2 segment that contains multiple access points. How important
it is then to reassign the "wide area mobility providing" anchor for
the remaining flows?

I would also keep the minimizing tunneling setup and amount of tunnel
state in the network as one design criteria.

> advertise both the prefixes, but the advertised prefixes should have the
> proper color to distinguish between those prefixes.

Agree.

> 3.) Network should set up tunneling for the old prefixes, to the prev
> anchor. It should deprecate the older prefixes.

See above about tunnels. These tunnels are essentially per prefix (or
even per address?) and a MN might soon be dragging quite few of those
along when it moves. I am a bit worried that the excess tunneling is
what we easily end up with..

> 4.) UE should use enhanced SAS schemes to use the new prefix for the new
> flows.

If you manage to deprecate an old prefix in 3) then new prefixes in 4) are
preferred automatically.. assuming the end host does existing RFC3484 to
any decent level.

> 5.) Over time, once the old flow die, the tunnel to the prev home is gone
> and the UE has optimized traffic flows.
> 
> This introduces simple changes into the mobility architecture and most
> semantics are already there in one form, or the other. The missing aspect is
> the type of Address and how we enable the UE to leverage that aspect and
> make Source Address selection and also for the network to manage those
> prefixes in a proper way. To that affect, I agree this is one important
> requirement.

- Jouni


> 
> 
> 
> -- 
> Regards
> Sri
> 
> 
> 
> 
> 
> On 3/19/12 3:16 AM, "[email protected]" <[email protected]>
> wrote:
> 
>> I think it is worth to work on solution providing more information on the 
>> type
>> of address. IMHO, it is a requirement for the UE to play with more than one 
>> IP
>> address and select the more appropriate source address according the
>> topological anchor point. DMM is clearly one use-case but, this feature is
>> also required for offload purpose in current centralized network based
>> mobility. The ongoing proposals for RA extensions, that allow to distinguish
>> anchored from local prefix, are interesting. Now, the UE shall have the
>> intelligence to make the source address selection according to prefix
>> properties; at least extensions to RFC3484 may be defined but more
>> sophisticated selection behavior may be required, typically, policies based
>> selection, e.g. offload policies.
>> 
>> Pierrick
> 

_______________________________________________
dmm mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/dmm

Reply via email to