Re: [netext] PMIP purists - is PMIP-NEMO impossible? draft-bernardos-netext-pmipv6-nemo-ps

["Seil Jeon" <[e-mail]>]

Hi Hesham,

Regards,

Seil

-----Original Message-----
From: Hesham Soliman [mailto:[e-mail]] 
Sent: Sunday, July 15, 2012 10:43 AM
To: Seil Jeon
Cc: [e-mail]
Subject: Re: [netext] PMIP purists - is PMIP-NEMO impossible?
draft-bernardos-netext-pmipv6-nemo-ps

>

>
>[SJ] Actually, there's no standard draft yet for PMIP-NEMO we're 
>discussing.
>Protocol designs for MR would be varied depending on the design 
>perspective.
>But one example could be as follows. Moving MAG would attach to PMIPv6 
>domain and have unique prefix assigned from a LMA. For attaching and 
>detaching mobile node, moving MAG should send PBU instead of MN 
>individually. However, in the case of mobile network's moving, PBU sent 
>by a MAG detecting mMAG's attachment to a LMA would be enough. Moving 
>MAG doesn't need to send anything for network moving event itself. 
>Following extended LMA binding cache entry can be an example.
>When the LMA receives PBU sent by the MAG regarding on moving MAG 
>attachment
>(MAG1 -> MAG2), it only changes tunnel endpoint of mMAG to MAG 2 and 
>doesn't need to touch MNs' entries belonging to mMAG.

=> Doesn't sound like something you would deploy on shared links, but I
guess that's a generic statement for PMIP.
Anyway, lets see when a draft is proposed.

[SJ] Yes, the given example was not based on shared link. For shared link
support, we may check another way.



>
>
>ID		Prefix		AR		M flag
>--------------------------------------------------------------------
>mMAG 1	Pref1::/64	MAG 2	no
>MN 1		Pref2::/64	mMAG 1	yes
>
>
>Actually, the above might be one example among many possible ways .. 
>and it's not my idea. FYI, you can see following.
>I just wanted to describe possibility of PMIPv6-NEMO support using 
>moving MAG.
>
>http://dx.doi.org/10.1109/MCOM.2009.4939291
>
>
>
>