Slashdot Mirror
The Military Aims to Develop 'Smart' & Secure WiFi
Krishna Dagli writes to mention a Network World article about a military project to create a self-configuring, secure wireless network. From the article: "Academic concepts such as artificial intelligence and Tim Berners-Lee's 'Semantic Web,' combined with technologies such as the Mobile Ad-hoc Network (MANET), cognitive radio, and peer-to-peer networking, would provide the nuts and bolts of such a network. Although the project is intended for soldiers in the field, the resulting advances could trickle down to end users. 'Military networks are going to converge as closely as we can to civil technologies,' says Preston Marshall, the program manager of DARPA's Advanced Technology Office."
reverse-engineered Goa'uld technology?
Please, for the good of Humanity, vote Obama.
I wouldn't be surprised if Google gets the contract. They practically hire every smart person in the world, so the military must be desperate to get this going.
The technology was first called the Online Occupation Infrastructure and Logistics network, but they thought the acronym would not be wise.
Well, when I was a youth I worked on the ARPA Network, a DARPA funded experiment in how networks recover from individual route failures. Well the technology grew up into the Internet. The US government wasn't pleased when they couldn't bomb away Saddam's communications network. It came out later that he used internet technology and that's why his network recovered so well. Now DARPA would like to do the same thing with inexpensive wireless devices. The technology is coming anyway, the genie is almost out of the bottle for good. Wirless networking is a disruptive technology that is inexpensive and flexible, I like it. I had a dream the other night about being a wireless guru and working with the south american rebels in the forest on their wireless network. Very exciting and dangerous. It would make a good movie.
The military doesn't want to rely on wireless technologies during warfare because they can be so easily jammed. All it takes is for someone to send noise on your frequency, and everything stops working.
Send email from the afterlife! Write your e-will at Dead Man's Switch.
"Military intelligence is a contradiction in terms."
Just think what they could do with WiFi.
Yes, you can dance to Radiohead.
They will probably use other frequencies. And I wonder how long it will take someone to build a device that blankets all the frequencies killing their network in an instant.
Wireless ad hoc nets have two major points of vulnerability: they are vulnerable to routing protocol attacks, and they consist of nodes with finite energy reserves.
I would disagree with the assertion in the article that current routing protocols are insufficient to handle MANETs. MANET routing protocols are slightly different (most are adaptations of traditional protocols), but if implemented correctly, they can support networks with very high rates of topology change... this has been supported by the literature for years now.
What the protocols are lacking is resistance from spoofing attacks that confound or exploit the "intelligence" of the adaptive routing protocols, and attacks on battery energy that coax nodes to use more energy or target and overwhelm key nodes. This has to be addresses in the lower layers as suggested by the article. So it's no surprise that the trend has been to develop "underlay" meshing protocols instead of traditional layer 3 routing schemes, because all of the security has to be built into layers 1 and 2 anyway on account of the fact that traffic can be easily sniffer or injected by passers by.
A networking protocol called "L2R" has been doing this for well over 10 years,
it is stable, very mature, evolved, and installed in dozens of places.
It was also shown to the US Military back in 2001 around the time of the trade center stuff.
They were interested, but couldn't understand it.
yes, that's right, the best the US DOD (at the time) had from their research facilities
couldn't understand the damn thing.
They even had a prof from the UC try and steal it and he made an RFC out of his understandings,
unfortuneately his understanding was so pathetically bad, it doesn't come close to working the same.
L2R is a proprietary, patented protocol, but because it's from a small, private group, it's never really caught on.
I seriously doubt anyone will ever figure it out at this point.
it's been over 10 years and still no one has even come close!!
I mean, really,come on guys, get a clue...
Don't assume that it is so easily jammed.
You assume that the wireless will be on a normal frequency. They could use spead spectrum or UWB. They could use light frequencies like infrared to carry the signal.
Or has that name already been used?
"Only two things are infinite, the universe and human stupidity, and I'm not sure about the former."
"wardriving"
the mods may say you posted flamebait, but to me it's a flame that warms my heart. rock on, brother! --chebucto
The F-22 has the ability to share radar, targeting, and other technical data with other F-22's. How is this done?
As if the government, or for that matter the military, could develop something complicated like a computer network.
.
.
.
Uhhhh wait....
So the design is oging to be great but someone is going to mess it up by setting the password to USA.
In the meantime, here are some reasons this is (even?) harder than it sounds.
"we'll probably keep the security work a little more isolated": offhand that sounds like it will never work. They're going to need security from the MAC layer upward and it's going to affect almost every decision they make.
It's a well studied problem, but one example of what you need to think through is the "hidden transmitter problem". It is possible to interfere with someone that you cannot hear.
TCP needs to be tweaked or replaced: packet loss doesn't mean the same thing in a chaotic RF world as it does on an Ethernet segment.
Media access policies need a lot of thought. Plenty of mesh algorithms have hit the wall when it came time to scale up the network. Again, as with routing, there are solutions but they haven't had as much time to grow up as TCP/IP has.
Hard problems make for fun projects. They should structure this so that they get something useful even if one of the ingredients fails: they're trying to do a lot of hard things at once. This kind of project is why we need DARPA: no sane private company would attempt this.
Why can't they use what is available now?
It seems to work fine for everyone else.
...already makes this stuff.
Microwave Data Systems
Informatus Technologicus
well, NATO jerks in general....used up quite a few expensive HARMs taking out microwave ovens in the war against the serbs. Great decoys. GO GO mil/indistrial complex, profits!
I though i read this a while ago on /. but maybe not. It's a DARPA project that uses wireless technology in anti-tank mines to "fix" themselves once breached. http://www.darpa.mil/ATO/programs/SHM/htmldemo.htm l
i thought i had lead poisoning until i stopped browsing at -1
Wireless: Either everyone can access it most of the time(open connection), or most people, not including you, can access it most of the time, and you can access it a small percentage of the time("Secure"). I don't know about you but this is my experience. I think it'll prove disasterous in the military.
The low-level details of these decentralized networks will be critical, but I am personally much more interested in the problems of applying this network to decentralized Command and Control in near-future urban combat. Squads will likely be temporary formations managed by decentralized algorithms. Dispersion of combat data will likely use gossip-like protocols and other ideas taken from modern P2P. The use of probabilistic flooding search will likely be more difficult because while networks like Gnutella have rapidly changing connectivity, they tend to retain the same basic statistics and topology. Because of rapid changes in a mobile network's connectivity statistics, the criticality threshold for probabilistic flooding as described by percolation theory will likely fluctuate as well, making it difficult to effectively operate near the transition region, where the number of flooded requests is minimized while still retaining good coverage.
Of course, for the wireless network to work correctly you must make sure that the dilythium reaction doesn't overload the quantum statis field or else the warp field will collapse, unless of course the power is diverted to the main deflector.
cause they have hired Al Gore to lead the project... based on his previous experience.... INVENTING THE INTERNET.
actually I am happy to see you, however that is in fact a banana in my pocket.
This kind of stuff has been around (for decades) on digital radio systems. Ericsson's is called EDACS, its a 9600 baud trunking control channel which is used to tell mobile radios which voice frequency to use. Modern 911 call centers have computers that dispatchers use to talk to fire/police/ambulance crews. There can be a lot of people wanting to talk all at once. Modern radios usually only have 20 or 30 channels to pick from, so they pick an electronic channel which is more like a database entry. If someone else is talking on a channel, and someone from another talk group wants to say something to everyone else on the same talk group, when they key up, the system automatically finds an empty frequency, tunes everones radios to that new frequency (without anyone lifting a finger) and everyone hears what is being said. Doing the same with wifi (best empty channel, least nodes/best performance using Dijkstras shortest path algorithm, and having things happen automagically is merely an extension on an old theme (although a cool one). There are already grid wifi networks too, so this technology actually fuses two existing technologies (one on a different platform) into something new.
MANET? It sounds like the government is developing an advanced gay porn site.
"Things are more moderner than before- bigger, and yet smaller- it's computers-- San Dimas High School football RULES!"
The problem with MANET is that of any matrix network: end points become bottlenecks. If there is only one node that has a connection with a gateway (quite normal in mobile environments) it ends up taking all the traffic of all the other nodes, which can become a bottleneck, a single point of failure and a target..
Not to mention that such excess activity will lean heavily on battery powered equipment (read: the most critical node is likely to fail first).
Good idea, but will take some work IMHO
With anonymity and encryption? GREAT! Let's ALL switch over now! I'll tell you what, you install the whole thing, give us real electronic voting while you're at it, and I'll LET you wiretap all my conversations and video chats starring thai strippers on pornotube.
Or how about getting senior members of government have a clue on what the internet even IS before you do one more god damned thing. Ted Stevens embarrassed us in front of the whole world a couple weeks ago.
Wait, series of tubes, pornotube.... I'm in!
rhY
I hold very few opinions. I hold information based on observation and fact. If you wish to disagree, please use facts.
Use this network, which essentially configures and extends itself where needed, as needed, to deliver HDTV, phone and Internet over IP. Wired, wireless or satellite - the network should be smart enough to use whatever means it has, but dumb enough to not care about what kind of traffic it routes, just that it does as good a job as possible with the available hardware. Automatically multi-link, it would route most of your p2p traffic through fibers while your VoIP goes wireless to your headset. Built-in authentication and encryption to keep your gadgets in touch and your data secure, even though you use someone elses hardware as well as let other use yours.
It's mesh networking, FON, cellphones, multicast and wimax, all the hype rolled into one big network. And no, we shall not call it Skynet.
Money for nothing, pix for free
2.4 Inter / Intra-Flight Data Link (IFDL)
Included in the CNI system is an Inter/Intra-Flight Data Link (IFDL) that allows all F-22s in a flight to share target and system data automatically and without radio calls. One of the original objectives for the F-22 was to increase the percentage of fighter pilots who make 'kills'.
With the IFDL, each pilot is free to operate more autonomously because, for example, the leader can tell at a glance what his wing man's fuel state is, his weapons remaining, and even the enemy aircraft he has targeted. This link also allows additional F-22 flights to be added to the net for multi-flight coordinated attack.
Caezar gave a talk on MANET at DEFCON. Don't know if the slides are available online, but he discussed a number of vulnerabilities in routing protocols as they apply to wireless, and also touched on some of the privacy issues. Very interesting talk... like most of the stuff at DEFCON. I'm actually surprised that this hit the media so fast.