Re: [RSN] Sleepy routers

I don't know if we need to address it directly, but I think that one of 
the really interesting questions is how to route over (one or more) L2s 
that have at least three orders of magnitude *configurable* tradeoff 
between two parameters of interest: data rate and lifetime.
There are WSN protocols in use today which can be configured to operate 
anywhere from 100% duty cycle down to well under 0.1%, on a link-by-link 
basis.
The simplest approach is to route with what is currently configured, and 
push the responsibility for configuration changes elsewhere.  Perhaps 
there are existing protocols that do this?  It seems like some of the 
existing traffic engineering work might be relevant. 
For example, I may know due to ND that I can talk directly to the final 
destination, but currently have no bandwidth configured (or not already 
spoken for) to do so.  Where does the tradeoff between QoS, user 
preferences for node lifetime, routing, and L2 duty cycle configuration 
happen?

I don't know how much, if any, of this fits in the scope of RoLL.  In 
any case, it is interesting to study.

ksjp

Ian Chakeres wrote:
> >From the proposed charter and name of the group I know that low-power
> and lossy are two important components. I was wondering whether these
> routers sleeping most (95%+) of the time should also be elevated to
> this status.
>
> What do you think? Is the fact that routers are sleeping a lot one of
> the core issues that needs to be solved in this WG?
>
> Ian
>
>
> _______________________________________________
> RSN mailing list
> RSN@ietf.org
> https://www1.ietf.org/mailman/listinfo/rsn
>   

_______________________________________________
RSN mailing list
RSN@ietf.org
https://www1.ietf.org/mailman/listinfo/rsn