Thread Links Date Links
Thread Prev Thread Next Thread Index Date Prev Date Next Date Index

Re: [802.21] FW: [DNA] Review of draft-ietf-dna-link-information-03.txt



Sure. But I still would like to understand why a simply, locally unique 
ID as I described below is insufficient for our purpose.

regards,
-Qiaobing

Guo-Qiang Wang wrote:

> Normally one link can be represented by two link termination point (with
> assigned address),
> so link ID is associated with two end points.
> 
> G.Q, Wang
> Nortel 
> 
> -----Original Message-----
> From: Qiaobing Xie [mailto:Qiaobing.Xie@MOTOROLA.COM] 
> Sent: Wednesday, May 24, 2006 1:18 PM
> To: STDS-802-21@LISTSERV.IEEE.ORG
> Subject: Re: [802.21] FW: [DNA] Review of
> draft-ietf-dna-link-information-03.txt
> 
> 
> 
>>If we want to create just an ID that does not help in 
>>routing/addressing, then I think a globally unique identifier with 
>>some cryptographic property would be the ideal, and we could consider 
>>compromises less than that.
> 
> 
> If 802.21 link ID is only for identification purpose by the MIHF that
> controls the link, then the ID can be assigned by the MIHF and only
> needs to be unique under the MIHF. E.g., MIHF (ID=12345) has three links
> (IDs=1, 2, 3) with their respective state=(up, down, going_down).
> 
> Here, since we have the concept of link_down, it kind of implies that we
> 
> are defining link as an endpoint (a downed link has no meaningful peer).
> 
> Of cause, an external entity can refer to a link by the MIHF-ID/link-ID 
> combination.
> 
> In the iab draft, it also says,
> 
> 
>>"... Each link endpoint has a unique link-layer identifier. "
> 
> 
> which seems to imply that iab's link-layer ID is also defined on the 
> endpoint.
> 
> regards,
> -Qiaobing
> 
>