Brad Boston | The future of ad hoc mobility

Widespread adoption of wireless networks is helping to spawn
another variation of Wi-Fi technology: mobile ad hoc networks.
MANETs comprise self-configuring and self-healing networks of
mobile routers that move and organize themselves arbitrarily.

Brad Boston, senior vice president of global government
solutions and corporate security programs at Cisco Systems,
recently met with Rutrell Yasin, Government Computer News'
senior technology editor, to discuss how MANETs are evolving.

Dave Buster, product marketing manager at Cisco's Global
Government Solutions unit, also joined the conversation.

GCN: Where do you see the most interest in MANETs in the
government at this time? Are defense and military services the main
users?

Brad Boston: Defense is out there driving it but also first
responders. Public safety has the same needs. So you think about an
environment that has been ravaged by a disaster ' whether
it's a disaster [such as the 2001 terrorist attacks] in
Manhattan or [Hurricane] Katrina down in New Orleans ' where
the entire fixed infrastructure gets wiped out in a certain area.
You need to be able very quickly to re-establish communications,
and what better way to do it than using ad hoc networking?

GCN:Is Cisco selling ad hoc networks to first
responders?

Boston:We haven't really pioneered it in the first
responder [market] yet. What we're doing in the first
responder area is primarily equipping them with mobile routers in
their vehicles, and cities like Manhattan are putting in a wireless
mesh infrastructure that the first responders can use. What we
think is the next step is to add the mobile ad hoc capabilities to
that deployment, which are basically some software changes.

GCN: How do you maintain an ad hoc network if its routers
come and go at random intervals?

Boston: That is one of the challenges: how the routers and
tables they maintain keep track of the different routing paths they
have. How do you go and allow that to be updated on a regular basis
in a very large network? That's one of our challenges. Today,
we can make it work fairly well with 50 to 100 nodes in the ad hoc
network. But if you get beyond that, you start getting degradation
in performance because they're spending too much [effort]
trying to reconfigure the stuff that is moving around.

GCN:Is it more than just adding bandwidth? Are there
standards-related issues that would help in maintaining ad hoc
networks?

Boston: Refining the way ad hoc networks, when they get
above a certain size, tables is probably the key breakthrough that
we're going to need.


Dave Buster: The academic community has been looking at
MANETs for a number of years now. But what we have noticed in the
market is a kind of separation of technology problems, and
different companies approach it from different directions.

We're trying to take a more holistic view. We're
looking at the radio problem, the way radios interfere with [one
another], and working with partners to build radios that work
better. We're taking information out of the radio and passing
it up into a router so that the router understands if there is
interference and can make choices, routing decisions, based on that
information.

Then we take that information and pass it up to the middleware
layers and build a construct that'[groups can use to
communicate] on these networks. And above that, we pass information
up into the application layer so a video source knows which video
codec to use, depending on the bandwidth that is available from the
network at that moment.

The point is, it is a top-tobottom kind of problem set that
needs to be solved at all layers simultaneously.

GCN: Are the present levels of wireless security
protocols and technology adequate for MANETs?

Boston:We're actually building in some additional
security capabilities as we [address] the problem '
especially in the defense area, where you may have people join your
network whom you want to allow to take advantage of the wireless
mesh you've formed, but you may not want them to have access
to all the other services that you're providing.

If you think about a coalition environment, you might want to
have your coalition partners use your network to reach back into
headquarters but not to share in the real-time chatter going back
and forth between the different assets that are part of the
network.

Buster:We've looked at several ways of actually
layering different security enclaves on top of the network so you
can have some individuals or nodes that are not trusted and they
are not allowed to participate. You have other individuals who are
semitrusted. They're allowed to transition the network, but
their traffic cannot terminate on any other node. They can only get
through the network.

Above that, we've built layers of security enclaves. For
much of what the military wants to do, they have their own
encryption schemes and we work with those. But for first
responders, they're looking more at commercial encryption
like [256-bit] Advanced Encryption Standard.

When we get finally down the road long term [toward industries
such as] mining, oil refineries, heavy construction, they want a
more lightweight security model that divides people up into
workgroups so they can work more effectively.

GCN: Can you give any real-world examples of ad hoc
network technology in use?

Boston: It is not being used on a large scale yet. But it is
being demanded very aggressively by the military. Think of the
[Warfighter Information Network- Tactical] project. It is basically
about mobile ad hoc networking. When you think about [the
Army's Joint Tactical Radio System], where they are trying to
build a router inside the radio, it is about mobile ad hoc
networking.

What we've been doing is demonstrating to [the military]
the commercial capability that exists today. We're now
working with a number of partners to get our big deployments.

We did a demo [in the spring] for a military organization that
showed multiple vehicles moving around in a theater of operation
with our routers integrated with the Harris military radio and
other technology, doing multiple video streams being shared between
all the videos back to the command post. We [also] had an airplane
join the network and allow it to be a node to help facilitate the
communications between the dispersed assets. There was also
satellite backup [in the] event that their radios lost sight of
each other.

GCN: What do you think is necessary for the technology to
take off in a big way?

Boston: One of the technology breakthroughs that [Buster]
alluded to was providing an interface between the radio and router.
What we need to do is get more of the radio vendors to build their
radios to that interface. Harris, maker of one of the military
radios and a big partner on the WIN-T program, has already
implemented that. We're working with other companies both in
the United States and [those that build] radios that the European
military uses.

On the first responders' side, we're trying to get
them to build fixed infrastructure first, and we'll add the
capabilities using the sophisticated radios they will deploy
there.

GCN: How do you see MANETs evolving? Boston:We
have to improve the scalability so more nodes can participate in
MANETs. The other thing to do is shrink the footprint of our
physical router that goes into vehicles so that it more easily can
be implanted in all types of vehicles that exist in the military
and public safety area. So we have small footprint routers today
that we need to improve horsepower on and add some other
features.

There are a number of software enhancements that we have in our
product road map that are requirements coming in from the United
States and other NATO military programs.