From majordom  Mon Feb  5 10:43:40 1996
Return-Path: <majordom>
Received: by itd.nrl.navy.mil (4.1/SMI-4.1)
	id AA25060; Mon, 5 Feb 96 10:43:40 EST
Received: from zia.epm.ORNL.GOV by itd.nrl.navy.mil (4.1/SMI-4.1)
	id AA25048; Mon, 5 Feb 96 10:43:33 EST
Received: (from sgb@localhost) by zia.epm.ORNL.GOV (8.6.12/8.6.12) id PAA19053 for mmnet@itd.nrl.navy.mil; Mon, 5 Feb 1996 15:43:34 GMT
From: "Steve Batsell" <sgb@zia.epm.ORNL.GOV>
Message-Id: <9602051043.ZM19051@zia.epm.ORNL.GOV>
Date: Mon, 5 Feb 1996 10:43:33 -0500
X-Mailer: Z-Mail (3.2.0 26oct94 MediaMail)
To: mmnet@itd.nrl.navy.mil
Subject: Open areas in setting up a working group
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Sender: owner-mmnet@itd.nrl.navy.mil
Precedence: bulk

mmnet mailing list:

As you all know, a BOF was held at the December IETF meeting in Dallas, to set
up a working group in mobile mesh networks. The idea of the working group is to
standardize both unicast and multicast routing protocols specificly design for
operation in a mobile mesh environment where both the hosts and routers move.

Joel Halpern, the routing area chair, has requested that before we can start a
working group we need to bound the scope of the problem  we will be addressing
in the working group.  This will include:

1. How rapid the motion can be and still expect the routing protocol to work?

< We covered several cases at Dallas including flooding>

2. How large an area must the protocol operate over?

3. How many nodes are expected to be in this area?

4. What type of hierarchy will be imposed on the system?

< The Army has a specific hierarchy they wish to operate within.>

While the orginal purpose of this working group was to specifically address the
military's need for a mobile mesh routing protocol standard both for unicast
and multicast, the working group should also address what commercial
applications would benefit/ require mobile mesh routing?  This is particularly
important as significant cost savings can be realized if commercial companies
in particular commercial routing companies were to implement these protocols in
a mobile router that DOD could purchase off the shelf.

As the discussion on the mailing list has been delayed, we have not made enough
progress to become a working group yet. If we feel we can sufficiently scope
the effort, Joel wopuld entertain holding a second BOF meeting at LA (March).
Alternately we could wait until the June meeting. I would like input from
everybody on this.

Getting this mailing list kicked off has been slowed for a number of reasons.
In January I moved from the Naval Research Lab to become the Networking
Research Group Leader at Oak Ridge National Laboratory. Please note my new
email address, sgb@ornl.gov.


--
Steve Batsell
Networking Research Group Leader
MS  6367, Bldg 6012
Oak Ridge National Laboratory
Oak Ridge, TN 37831
423-574-7401 (phone)
423-574-0680 (Fax)
sgb@ornl.gov

From majordom  Mon Feb  5 12:45:22 1996
Return-Path: <majordom>
Received: by itd.nrl.navy.mil (4.1/SMI-4.1)
	id AA03874; Mon, 5 Feb 96 12:45:22 EST
Received: from [132.250.92.68] (gumbo.itd.nrl.navy.mil) by itd.nrl.navy.mil (4.1/SMI-4.1)
	id AA03852; Mon, 5 Feb 96 12:45:11 EST
Date: Mon, 5 Feb 96 12:45:10 EST
Message-Id: <9602051745.AA03852@itd.nrl.navy.mil>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
To: "Steve Batsell" <sgb@zia.epm.ORNL.GOV>
From: macker@itd.nrl.navy.mil (Joseph P. Macker)
Subject: Re: Open areas in setting up a working group
Cc: mmnet@itd.nrl.navy.mil
Sender: owner-mmnet@itd.nrl.navy.mil
Precedence: bulk

>
>Joel Halpern, the routing area chair, has requested that before we can start a
>working group we need to bound the scope of the problem  we will be addressing
>in the working group.  This will include:
>
>1. How rapid the motion can be and still expect the routing protocol to work?
>
>< We covered several cases at Dallas including flooding>
>As the discussion on the mailing list has been delayed, we have not made enough
>progress to become a working group yet. If we feel we can sufficiently scope
>the effort, Joel wopuld entertain holding a second BOF meeting at LA (March).
>Alternately we could wait until the June meeting. I would like input from
>everybody on this.

Steve/mmnet group, 

My initial vote is to hold off until June so that we may further
discuss/define the focus of this effort.  I am willing to meet in LA if we
decide it is a good thing. If we postpone until June, we can use the time
to kick up more active technical exchange on this mailing list.

-Joe

    _/      _/ _/_/_/_/  _/       Joseph P. Macker, Code 5544
   _/_/    _/ _/     _/ _/        Naval Research Laboratory
  _/  _/  _/ _/_/_/_/  _/         Center for High Assurance Computer Systems
 _/    _/_/ _/   _/   _/          macker@itd.nrl.navy.mil
_/      _/ _/     _/ _/_/_/_/     http://torrant.itd.nrl.navy.mil/macker.html



From majordom  Mon Feb  5 16:13:12 1996
Return-Path: <majordom>
Received: by itd.nrl.navy.mil (4.1/SMI-4.1)
	id AA19626; Mon, 5 Feb 96 16:13:12 EST
Received: from newra.src.umd.edu by itd.nrl.navy.mil (4.1/SMI-4.1)
	id AA19613; Mon, 5 Feb 96 16:13:03 EST
Received: from oxygen.src.umd.edu (oxygen.src.umd.edu [128.8.111.61]) by newra.src.umd.edu (8.7.3/8.7) with ESMTP id QAA23569 for <mmnet@itd.nrl.navy.mil>; Mon, 5 Feb 1996 16:13:16 -0500 (EST)
From: Mathew Scott Corson <corson@isr.umd.edu>
Received: (from corson@localhost) by oxygen.src.umd.edu (8.7.3/8.7) id QAA00955 for mmnet@itd.nrl.navy.mil; Mon, 5 Feb 1996 16:13:01 -0500 (EST)
Date: Mon, 5 Feb 1996 16:13:01 -0500 (EST)
Message-Id: <199602052113.QAA00955@oxygen.src.umd.edu>
To: mmnet@itd.nrl.navy.mil
Subject: scope
Sender: owner-mmnet@itd.nrl.navy.mil
Precedence: bulk

Everyone,

Regarding Joe's comment...

> 
> My initial vote is to hold off until June so that we may further
> discuss/define the focus of this effort.  I am willing to meet in LA if we
> decide it is a good thing. If we postpone until June, we can use the time
> to kick up more active technical exchange on this mailing list.
> 

...postponement is probably best in my opinion.  I'll also be in LA if 
anyone wants to meet and discuss things offline.

Regarding what Steve said...

> 
> Joel Halpern, the routing area chair, has requested that before we can start a
> working group we need to bound the scope of the problem  we will be addressing
> in the working group.  This will include:
> 
> 1. How rapid the motion can be and still expect the routing protocol to work?
> 
> < We covered several cases at Dallas including flooding>
> 
> 2. How large an area must the protocol operate over?
> 
> 3. How many nodes are expected to be in this area?
> 
> 4. What type of hierarchy will be imposed on the system?
> 
> < The Army has a specific hierarchy they wish to operate within.>
> 

...the group could use some input from people with specific numbers in mind
regarding the preceding items for particular applications.

Personally, I feel this is not the best way to specify the group's scope.
I think any numbers we come up with should be viewed as food for thought, 
but not as absolute requirements.  Routing protocol design will be driven 
by the need to operate within highly dynamic topologies over congested 
links with potentially differing link level access schemes.  The best 
(i.e. most efficient in terms of complexity) solution will be adopted 
and the numbers, which can truthfully become known only after simulation, 
will reflect our best design.  In retrospect, this will tell us whether 
the numbers were realistic or not with given technology.

I think it better if we first define our scope "qualitatively" in 
architectural terms.  This is basically the theme of the draft RFC which
was authored.  The group's preliminary charter defined a mobile mesh network
(mmnet) essentially as an ad hoc network consisting of mobile routers 
connected by wireless links, typically characterized by a dynamic topology 
and congested operation. 

At the BOF, Joe Cleveland gave a presentation on the usefulness of
flooding in this situation.  Also, Dave Johnson mentioned the possibility 
of using dynamic source routing.

I gave a presentation which:
o motivated the need for multipath routing for congestion
  avoidance in mobile networks;
o motivated the need for highly adaptive, light-weight,
  routing to quickly react to connectivity changes
o gave a preliminary set of broad design assumptions:
	- flat routing architecture (intradomain)
	- decoupling of routing and multiple access (necessary for operation
		across multiple subnetwork technologies)
	- multipath routing desired when possible
	- quick, non-optimal route construction necessary
	- limited use of flooding

Each of the three approaches mentioned at the BOF (flooding, dynamic source 
routing and highly adaptive routing) have merits and limitations with 
respect to one another in various situations.  

At present, my feeling is that a "hybrid" protocol which combines the 
best features of these three approaches may be a good solution.

Does anyone have any comments on this or other architectural issues, or 
have any specific numbers for the group's scope mentioned previously?

-Scott


From majordom  Sun Feb 11 11:31:49 1996
Return-Path: <majordom>
Received: by itd.nrl.navy.mil (4.1/SMI-4.1)
	id AA09630; Sun, 11 Feb 96 11:31:49 EST
Received: from [132.250.92.68] (gumbo.itd.nrl.navy.mil) by itd.nrl.navy.mil (4.1/SMI-4.1)
	id AA09580; Sun, 11 Feb 96 11:31:38 EST
Date: Sun, 11 Feb 96 11:31:37 EST
Message-Id: <9602111631.AA09580@itd.nrl.navy.mil>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
To: mmnet@itd.nrl.navy.mil
From: macker@itd.nrl.navy.mil (Joseph P. Macker)
Subject: Archive Retrieval
Sender: owner-mmnet@itd.nrl.navy.mil
Precedence: bulk

Hello group:

Mailing list members can now get an archive history by sending mail to
majordomo@itd.nrl.navy.mil with the body "get mmnet mmnet.archive".

The mars site is not receiving the latest archive.
-Joe

    _/      _/ _/_/_/_/  _/       Joseph P. Macker, Code 5544
   _/_/    _/ _/     _/ _/        Naval Research Laboratory
  _/  _/  _/ _/_/_/_/  _/         Center for High Assurance Computer Systems
 _/    _/_/ _/   _/   _/          macker@itd.nrl.navy.mil
_/      _/ _/     _/ _/_/_/_/     http://tonnant.itd.nrl.navy.mil/macker.html



