Re: [Ospf-manet] URL for MPR-extension software

Emmanuel Baccelli <> Fri, 22 December 2006 17:17 UTC

Received: from [] ( by with esmtp (Exim 4.43) id 1Gxo1m-00076t-2T; Fri, 22 Dec 2006 12:17:50 -0500
Received: from [] ( by with esmtp (Exim 4.43) id 1Gxo1k-00076o-Mp for; Fri, 22 Dec 2006 12:17:48 -0500
Received: from ([]) by with esmtp (Exim 4.43) id 1Gxo1b-0004Yc-Su for; Fri, 22 Dec 2006 12:17:48 -0500
Received: from [] ( []) (authenticated bits=0) by (8.13.6/8.13.6) with ESMTP id kBMHHY2W000662 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO) for <>; Fri, 22 Dec 2006 18:17:34 +0100
Message-ID: <>
Date: Fri, 22 Dec 2006 18:17:33 +0100
From: Emmanuel Baccelli <>
User-Agent: Thunderbird (Macintosh/20061207)
MIME-Version: 1.0
Subject: Re: [Ospf-manet] URL for MPR-extension software
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
X-Miltered: at discorde with ID 458C132E.000 by Joe's j-chkmail (http://j-chkmail . ensmp . fr)!
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 39bd8f8cbb76cae18b7e23f7cf6b2b9f
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Discussions of OSPFv3 extensions supporting MANET <>
List-Unsubscribe: <>, <>
List-Archive: <>
List-Post: <>
List-Help: <>
List-Subscribe: <>, <>

Dear Richard, all,

Just to clarify, the MPR proposal offers topology reduction, adjacency
reduction, and uses multicast retransmissions. It was initially
implemented based on the OR code in order to produce something which is
compliant with the protocol and operational within the simulator in a
timely fashion. Other than adjacency reduction, which is optional, we
believe that the simulation code complies to the specification, however
if you can point out specific issues (and even have a patch) we welcome
and appreciate the input from you as well as from the rest of the WG. We
believe that it is in everybody's interest to progress as decided at the
San Diego meeting to get the three current proposals matured and forward 
to experimental, and as you, we are working hard towards that goal.

To address your specific technical concerns from previous emails, it 
seems you point out that topology reduction is a key component. We 
agree. Both MPR and MDR solutions feature mechanisms for this purpose. 
The MPR-OSPF protocol implementation for GTnet that has been made
available and tested shows that MPR topology reduction is simple, 
efficient, and fits well with OSPF requirements (such as symetric link 
advertizement, and stability with regards to received LSAs).
We understand that extensions to the MDR proposal with a new topology
reduction scheme are envisioned in the future. We are looking forward to 

Synchronization is also an important issue and, as you point out, the 
two different ways (direct and indirect synchronization) feature 
different advantages. The MPR adjacencies seem to offer a good 
compromize - in particular the use of multicast greatly reduces the 
weight of adjacencies in LSA transmission, and simulations seem to 
confirm this. We are currently in the process of double-checking the 
code that implements this, which will, of course, shortly be made available.

With the best wishes for a happy holiday season to you and yours,


Ospf-manet mailing list