[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
RE: [manet] AODV Question
IMHO ... John's application is why simple QoS metrics (e.g, "node energy status" or "link stability") might should be included in the next round of MANET WG proposals. Similar problems exist in military sensor networks.
--Pete Sholander
Atlanta, GA
>>> "Manfredi, Albert E" <albert.e.manfredi@boeing.com> 11/17/03 11:22AM >>>
Hi John,
Seems to me that what you describe ought to work just fine with any IGP. Wired networks also have routers that go up and down, after all.
I think this is only a matter of degree. If the infrastructure is not meant to be ad hoc, i.e. the density of nodes that behave as routers is roughly what it needs to be to provide reliable connectivity, such that flooding of hello messages will not overwhelm an overly dense set of routers, then you don't need to rely one special enhancements of standard IGPs.
All you're describing, I think, is a typical internet in which routers happen to have RF links?
Bert
> -----Original Message-----
> From: John Mullen [mailto:jomullen@nmsu.edu]
> Sent: Monday, November 17, 2003 11:07 AM
> To: manet@ulfius.com; manet@ietf.org
> Subject: RE: [manet] AODV Question
>
>
> Hi Charlie,
>
> Well, I can see one application of MANET technology in a situation in
> which the nodes do not move. Consider the Dine' (Navajo) nation which
> covers thousands of square miles in the southwestern US.
> Connecting the
> many isolated communities with wire or fiber does not seem to be an
> option. Furthermore, the Dine' would prefer not to use the
> ideal places
> for antennas because of religious and other reasons. Every
> time you dig
> a hole on the reservation, you need an archeologist. Other American
> Indian jurisdictions are checkerboarded. That is, there are
> parcels of
> fee simple land owned by non-members of the Tribe interspersed
> throughout the area. In those cases, right of way is a big problem.
>
> It is not just that putting up telephone poles is a problem, but also
> that a common electrical supply is often not available. That is, any
> repeaters would have to be self-powered.
>
> Now, imagine a network of nodes that are powered by
> batteries, which in
> turn, are recharged by solar energy. Also, assume the net is dense
> enough that many nodes can go to standby while the others
> take the load.
> However, as the evening wears on, the active nodes, one by one, will
> have to shut down due to low power. Others could, at any
> time, have to
> shut down due to malfunctions (we a talking of many nodes in a very
> hostile environment). As each does so, one or more inactive
> nodes wakes
> up to take the load.
>
> Such a net would have no mobile nodes, but would require frequent
> automatic reconfiguration. It wouldn't be entirely a MANET,
> since each
> node could know where every other node was, but would require a degree
> of automatic reconfiguration that ordinary protocols would not support
> due to the frequent changes in the pattern of available
> nodes. I would
> think that at least part of a MANET protocol would be
> helpful. Also, if
> such a protocol were to be used, mobile nodes could be more easily
> accommodated as they move throughout the communications area.
>
> Just a thought.
>
> John Mullen
>
> -----Original Message-----
> From: manet-admin@ietf.org [mailto:manet-admin@ietf.org] On Behalf Of
> Charles E. Perkins
> Sent: Monday, November 17, 2003 8:27 AM
> To: DANIEL BYRNE
> Cc: 'Erik Nordström'; manet@ulfius.com; manet@ietf.org
> Subject: Re: [manet] AODV Question
>
>
>
> Hello folks,
>
> One or two comments below:
>
> DANIEL BYRNE wrote:
>
> > Yes we did, but everyone here is now saying that AODV doesn't work
> > well over 50 or more hops.
>
> I can believe that the performance would suffer
> greatly, depending on the mobility of the intermediate
> nodes. If the nodes didn't move, then I wonder what
> would go wrong?
>
> Regarding the idea about replacing land lines,
> I guess that any routing protocol would work,
> even manual configuration. To overcome the
> hop limit, you could do encapsulations.
>
> It doesn't seem to be a close match for the
> technology provided for use with mobile ad
> hoc networks since:
> 1) the nodes aren't mobile, and
> 2) the network isn't ad hoc.
>
> Regards,
> Charlie P.
>
> _______________________________________________
> manet mailing list
> manet@ietf.org
> https://www1.ietf.org/mailman/listinfo/manet
>
>
> _______________________________________________
> manet mailing list
> manet@ietf.org
> https://www1.ietf.org/mailman/listinfo/manet
>
_______________________________________________
manet mailing list
manet@ietf.org
https://www1.ietf.org/mailman/listinfo/manet
_______________________________________________
manet mailing list
manet@ietf.org
https://www1.ietf.org/mailman/listinfo/manet