Re: [DNA-BOF] Re: [mobile-ip] Announce: BoF Proposal: DetectingNetwork Attachment

[Greg Daley <greg.daley@eng.monash.edu.au>]

Hi JinHyeock,

JinHyeock Choi wrote:
> Dear Thomas
> 
> Thanks for your comments. It helped to clarify my thoughts. 
> 
> 
>>>The notion of 'Detecting Network Attachment' is not clear to me. 
>>>What is the difference with 'Movement Detection'?
>>
>>If you disconnect from one network and reconnect to another one,
>>you've just moved. Thus, detecting movement is closely tied to
>>determining that you have a new connection (or potential connection)
>>to a network (or have become disconnected from an existing one) in the
>>first place.
>>
>>Thus, "movement detection" logical builds on "detection" in general.
>>You can't determine whether you've moved before determining that
>>you've attached or detached.
> 
> 
> Correct me if I am wrong. I express their difference like below (though
> it needs more elaboration and clarification. 
> 
> For "Movement Detection", a mobile node performs the following tasks
> 1) Checking the (partial) reachability of current AR 
> 2) Checking the validity of current CoA
> 3) The discovery of new AR with necessary information, for example 
>     on-link prefix.
> 
> For "Detecting Network Attachment", a node performs the following task,
> 1) Checking to which routers it is attached.
> 2) Checking whether it has valid IP address.
> 3) If necessary, getting a RA with with necessary information, for example 
>     on-line prefix. 
>  
> Correct?

I think this is a fair guess at what a v6 node
might wish to do (in general).

The differences between MD and DNA as described
here may be a bit subtle though.  We'll definitely
need to determine what the problem space is,
when we get into the work.

Existing requirements will play a big role in
this.

Greg