Hi,

Wired or Wireless is not really the point.
I think we should rather draw a line between the Lossy and the Very Reliable 
Networks (which is often tight to the power/size/cost constraints).

I think a routing protocol can run over wired or wireless links, as soon as it 
is a route-over protocol.
The real point is the set of requirements taken into considerations for the 
design of the routing protocol.

For instance, we use RPL over PLC (wired) and 802.15.4 (wireless) because they 
both fall in the LLN category (which is indeed not restricted to wireless 
links).

To my mind it is clear that we have to use a route-over protocol if we consider 
multiple technologies (which I believe is a correct assumption in a home).

Some examples to illustrate my opinion :

Lossy Wired : Some PLC technologies (See AMI technologies).
Reliable Wired : Ethernet
Lossy Wireless : 802.15.4
Reliable Wireless : Wifi

Finally, I really agree with a previous remark on the list who states that the 
single magic routing protocol is hard to find.
We may use some routing protocols, depending on the type of devices we think 
will be used in the home.

A remark about this statement :

"I think OSPF or IS-IS (or even EIGRP) can
be adapted to the low power/lossy network situation, it just takes some
thought and care."

This doesn't seems to be compliant with the protocol survey which has been done 
by the ROLL WG :

See Figure 1 here : 
http://tools.ietf.org/html/draft-ietf-roll-protocols-survey-07#page-14

Best,

Cédric.

Le 3 oct. 2011 à 22:53, Ulrich Herberg a écrit :

Hi,

On Mon, Oct 3, 2011 at 1:45 PM, Russ White <ru...@riw.us<mailto:ru...@riw.us>> 
wrote:

Can you also say the opposite that rotocols developed for wireless work
equally well also for wired environments ? If so let me ask why do we
need both classes of routing protocols ?

I have deployed OLSRv2 both on Ethernet and on wireless environments,
without problems. I do not see a reason why OLSRv2 could not run in
mixed environments.


I would argue that the OSPF MANET extensions would work just as well for
IS-IS or OSPF in both wirelss and wired networks.

The OSPF MANET extensions would certainly be a valid alternative to
consider. Which one performs better or is more suited (OLSRv2 /
OSPF+MANET / ...) -- I don't know, that would be an interesting
comparison.


If not I am not sure homenet should focus on wireless router to router
communication at all. Yes I am in favor of wired env with fiber or with
nice cat 7 wiring between my routers.

I am as well, but I think we should allow for both with a single
protocol as much as possible.

I agree. Not everyone likes long cables in a home environment. And
having a single protocol would be indeed better.

I think OSPF or IS-IS (or even EIGRP) can
be adapted to the low power/lossy network situation, it just takes some
thought and care.

We have invested that thought and care in MANET for years now. I think
it could be worth investing the suitability of the protocols developed
in MANET (as well as the OSPF MANET extensions), before reinventing
the wheel.

Ulrich



Russ

_______________________________________________
homenet mailing list
homenet@ietf.org<mailto:homenet@ietf.org>
https://www.ietf.org/mailman/listinfo/homenet

_______________________________________________
homenet mailing list
homenet@ietf.org<mailto:homenet@ietf.org>
https://www.ietf.org/mailman/listinfo/homenet


_______________________________________________
homenet mailing list
homenet@ietf.org
https://www.ietf.org/mailman/listinfo/homenet

Reply via email to