Re: [DNA] Considerations for DNA Schemes with multiple Interface andLayer 2 Technologies

[JinHyeock Choi <jinchoe@diffeo.com>]

Dear Yong-Geun

Thanks for your reply. 

>  Because I am not a specialist about CDMA, I can be wrong in some points.
>  As my understanding (with discussion with other CDMA specialists), 
> it is possible for a host to distinguish the IP packets from the old
> link-layer connection FROM the ones
> from the new link-layer connection. 

Glad & relieved to hear that. 

> The reason is that a host in CDMA can
> have two link-layer connections 
> during handover between base stations. One link-layer connection is for an
> old base station, the other link-layer
> connection is for a new base station. When handover happens and by using
> make-before-break, a host can have 
> the ability to distinguish the IP packets. 

I am confused. Why does a host need 'make-before-break' 
technology to distinguish the IP packets?

In case there is no 'make-before-break', only one link-layer connection 
is allowed per interface. Hence, all the packets come from the same 
link-layer connection. No need to distinguish packets at all.  

> (Actually, current IPv4 based
> CDMA packet service in Korea, unfortunately
> only supports break-before-make(hard handover) because there are no needs
> for make-before-break(soft handover) )
>  So, if this situation is valuable for DNA schemes,  we can ask CDMA
> operators to support make-before-break in 
> IPv6 based CDMA packet service. 

Allow me to present it more logically. 

First,  
If CDMA supports 'make-before-break', it's possible for an interface 
to establish multiple link-layer connections at the same time. 

Second,  
If so, it's difficult for an interface to detect the identity of the currently 
attached link, because there may be more than one currently attached links. 

Third,  
In this case, it would help DNA scheme if an interface can distinguish 
the packets from one link-layer connection from another. 

Purely DNA point of view, it would be better if CDMA doesn't support 
'make-before-break' at all. It would make DNA much simpler. 
 
Best Regards

JinHyeock