Re: The argument for writing a general purpose NAT for IPv6

[EricLKlein@softhome.net]

james woodyatt writes: 

> On Apr 18, 2007, at 00:51, EricLKlein@softhome.net wrote:
> > To Paraphrase Alex Lightman:
> > When a mobile device has a private address, it receives an endless  
> > series of "keep alives". The carrier's NAT wants to assign the  private 
> > number to another device, and so it keeps "checking the  roast", but this 
> > burns up the battery. With an IPv6 address once  the connection is made 
> > there is no need to keep "checking the  roast" every few seconds. -
> > http://www.usipv6.com/6sense/2004/oct/october03.htm
> > This seems worth keeping NAT out of IPv6 for mobile devices alone  in 
> > addition to the reasons that were in our document.
> Won't the keep-alives need to be sent from the mobile device to keep  the 
> *filter* state from timing out whether there is NAT in the loop  or not?  
> It seems to me this is just a fact of life that mobile  wireless devices 
> are going to have to endure, simply as a result of  our insistence that 
> gateways implement asymmetric stateful packet  filters in factory default 
> configurations.
James, 

It is my understanding that IPv6 based mobile devices will need to do keep 
alives at least half as many times as IPv4 NAT based devices. That said, I 
am not sure what the affect that stateful packet filters will have on the 
flow and if they will be implemented at the device or in the network 
upstream of the mobile device to save radio bandwidth. 

Maybe one of the guys from Nokia (who did the testing of battery life) can 
answer this.