[ndnSIM] Longest prefix match in FIB
alexander.afanasyev at ucla.edu
Mon Apr 14 11:51:17 PDT 2014
I think you're right. FIB works entirely on names and does not really need to know anything about the Interest. I think the original idea was to match PIT implementation (where PIT entry is looked up either with Data or Interest packet), but I see now that it is not entirely correct.
If you make a patch that adds a new LPM method that uses just a name (I would prefer keeping the existing method for now), then I would be happy to merge this into the master.
On Apr 10, 2014, at 7:08 AM, Yaoqing Liu - liu <liu at clarkson.edu> wrote:
> Hi Alex,
> I'm working with Dr. Lan Wang from the University of Memphis about the forwarding hint solution. We noticed that your only implemented one Longest Prefix Match (LPM) function by interest, we have two questions here:
> why don't just use a prefix for LPM, instead of an interest (In the function, actually you used just the interest name)?
> You left a note in the function "@todo use predicate to search with exclude filters", what other fields of an interest and for what purpose will be used for FIB lookup? Could you show an example? In our opinion, LPM by the interest name is good enough for FIB lookup.
> Yaoqing Liu, Ph.D.
> Assistant Professor
> Computer Science
> Clarkson University
> Potsdam NY 13699
> Office: Science Center 381
> Phone: (315)-268-2374
> Email: liu at clarkson.edu
> Web: http://people.clarkson.edu/~liu/
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the ndnSIM