[Cu-wireless] [mlsichit@unity.ncsu.edu: [manet] Re:Autoconfiguration protocols]

[David Young]

Important ideas for address assignment here. I remember that in some
early meetings, we discussed similar ideas. I wonder what keeps the
addresses from concentrating at and near the network's "original" node?
I wonder when I will read the paper to find out? =)

Dave

----- Forwarded message from Mihail L Sichitiu <mlsichit at unity.ncsu.edu> -----

Date: Tue, 11 Feb 2003 14:10:28 -0500 (EST)
From: Mihail L Sichitiu <mlsichit at unity.ncsu.edu>
To: manet at ietf.org
Subject: [manet] Re:Autoconfiguration protocols


Hi,

I used to briefly work at autoconfiguration problem until I came to know
that "my approach" was recently published in an MS thesis on the subject
(at UT. Dallas - see link below). I have nothing to do with the thesis,
but I thought it's relevant.

I think that the approach in the thesis (which draws on an unpolished idea
of an U. of Maryland student) is significantly more scalable (O(m)
messages instead of O(nm) for m new joins in an n node ad-hoc network)
than existing approaches, and deserves your attention. I honestly doubt
that it can be done much better than that, but I'm looking forward to see
a better solution).

http://www.utdallas.edu/~mansi/

If you want to get a look at the U. of Maryland initial papers see
references [12] and [13] in the thesis. I was unable to find them online,
but I got a copy from the library.

The main idea of the thesis (and original papers) is simple: each node in
the ad-hoc network acts like a DHCP server that has besides its IP address
a range of IP addresses that can offer. When a new node joins an already
existing node it offers an IP address and half of its IP addresses in its
care. The thesis takes care of other details like network merging and
partitioning, and address reclaiming.

Best regards,
Mihail Sichitiu

_______________________________________________________
Mihail L. Sichitiu
Assistant Professor,              Phone: (919) 515 7348
Dept. of ECE, Campus Box 7911     Fax:   (919) 515 5523
NC State University
Raleigh, NC 27695-7911   http://www4.ncsu.edu/~mlsichit
_______________________________________________________
_______________________________________________
manet mailing list
manet at ietf.org
https://www1.ietf.org/mailman/listinfo/manet

----- End forwarded message -----

-- 
David Young             OJC Technologies
dyoung at ojctech.com      Engineering from the Right Brain
                        Urbana, IL * (217) 278-3933