Re: [DNA] Revised FRD (Fast Router Discovery) I-D

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

Hi Syam, 

----- Original Message -----
[cut]
> > Also, you definitely need modified routers, but I guess
> > that's handled by TSLLAO.
> 
> Yes. Also TSLLAO is not required all the time, in such cases 
> the router would send multicast RA.

I think that it's OK for the router to send a multicast RA,
but it may be necessary to get a fast response from the router
to use TSLLAO in the RS, otherwise the response is at the 
mercy of the MIN_DELAY_BETWEEN_RAS interval.

[cut]
> > 
> > It's really not possible to tell how much simpler it is, though.
> > The APs need to track a list of all routers, some of which may
> > not be trusted at L3 (Do they do SEND?).  
> 
> APs either can be manually configured or can be learnt from RAs.
> I am thinking that APs/Base Stations are more trustable nodes.
> Are you saying the APs may be trustable at L2 but not at L3?
> I do not know much about SEND, but I think it can be used if
> required.

I was thinking the other way around: that the APs may need
to be able to trust the routers, in order to select appropriate
routers to solicit.

Also I don't think APs are trustable, except if there's a particular
reason to trust them - which could be described if need be (I was
concentrating on trust of ARs though so this may not be necessary).


Greg