Military To Spend $42M To Build Advanced Network Control 102
coondoggie writes "BBN, which was bought by defense giant Raytheon today, got almost $11 million to help build self-configuring network technology that would identify traffic, let the network infrastructure prioritize it down to the end user, reallocate bandwidth between users or classes of users, and automatically make quality-of-service decisions.
The advanced network technology is being developed by Defense Advanced Research Projects Agency (DARPA) and will include support for features like 32 levels of network traffic prioritization that will let data with a higher priority will be handled more expeditiously than traffic with a lower priority."
Dose it (Score:3, Funny)
Re: (Score:2, Funny)
Re:Dose it (Score:5, Funny)
Early experiments using the STFU protocol showed that network traffic went to zero. While this had positive cost impact, for example because you could omit all those costly cables without further harm, it was finally concluded that data rates above zero had enough advantages to offset those costs.
Uhm (Score:2, Insightful)
And what exactly is low priority traffic?
Re:Uhm (Score:5, Funny)
Your post.
Re:Uhm (Score:4, Interesting)
Re:Uhm (Score:5, Funny)
Well, every manager knows that power point slides have always the highest priority. Fuck those missiles.
Re: (Score:2, Informative)
Fuck those missiles.
I believe the expression is "Damn the torpedoes!"
Re: (Score:1)
That is soooo 2006... :)
Re: (Score:1)
Re: (Score:2)
"whereas an Alert for an incoming cruise missile to the Command and Control Systems might be considered slightly more high priority."
Those shouldn't even be on the same physical network.
Re:Uhm (Score:5, Insightful)
The lowest bidder.
Re: (Score:2)
Whereas you always buy from the highest bidder, right?
Re: (Score:2)
Since the context is someone paying you for a limited resource, then yes I usually would...
Re: (Score:2)
I have a product to sell you for $infinity. I can guarantee that you won't find a better price.
Re: (Score:2)
That I would likely take the highest bid when selling something says nothing about what I would do when buying something.
So I fail to see the relevance.
Re: (Score:1)
Re: (Score:1)
Why would a journalist be posting a non-desirable story inside the military's internal network? I'm pretty sure this will be applied to the military's internal classified networks for better transferring important command and control information at a higher priority than non mission and time critical data.
This isn't likely to apply to the internet as a whole, just their own internal networks. And I'm sorry, but in an operational environment where people can die if the right information doesn't get there on
Wow (Score:4, Insightful)
$11M to reimplement IPv6 QOS. I suppose it's a bit more advanced since it makes QOS determination based on users or groups, but that doesn't seem that difficult.
Consider me unimpressed.
Re:Wow (Score:4, Funny)
Someone else with no experience doing massive implementations of new infrastructure spouting off.
Consider me unimpressed.
Re: (Score:1)
Re: (Score:1)
Oh! and I thought it was al gore.... had it all wrong ;)
Re:Wow (Score:5, Insightful)
$11M to reimplement IPv6 QOS. I suppose it's a bit more advanced since it makes QOS determination based on users or groups, but that doesn't seem that difficult.
Consider me unimpressed.
Dude, there might be strategic/tactical decisions for deciding explicitly not to use IPv6. Notice that I'm not saying that those strategic/tactical decisions are necessarily valid for long-term maintenance, extensibility or external compatibility (the later of which might even be undesirable from a strategic/tactical POV.)
The road of technical divergence can either take you to innovation or to a complete technical fiasco. That fork is many times not only dependent on technical merits alone. Besides, iirc, IPv6 QOS is still as of yet to be developed (not a criticism mind you). It supports only 7 priority levels whereas the proposed technology will support 32 levels. A typical military subnet, with stationary and mobile units, all of them plugged and receiving feeds from a bunch of disparate devices might never need more than 7 or 8, but as you start plugging those nets together, you can (and will) easily require a finer priority granularity than that.
Add to that the ability to determine priority by user or groups, and the problem cannot be dismissed as "meh, should not be that difficult." There might be other defense-specific requirements that we might not know (.ie. limiting jumbopackets by priority or origin.)
Besides, this is being researched by DARPA, the harbinger of ARPANET and MQ-Predator, not some 2009 rendition of kozmo.com.
I know that here on /. we like to fling turds at the government's white elephants, but c'mon. There must likely be be good technical/domain-specific reasons (or at least good enough) for an entity like DARPA to perform research on it, reasons beyond the ones that might impress you.
Re: (Score:1)
Assigning priority by user *cannot* be a hard problem, because anything you'd do to make it more than trivial would make it impossible.
The options are essentially: User == source address or: User is embedded in packets because if you don't either know who the user is explicitly or provide some sort of mapping from "user" to "something discrete already in the packet" there's no way to determine the user for arbitrary traffic.
And once you've established which user sent a packet assigning the priority is just a table lookup, which is something routers are already pretty good at.
Well, if you paint all user-mapping problems in such a generic way, then obviously that problem *cannot* be that hard.
Determining who the user is when you know a-priori that it is at the network layer (user == address) or at the transport/application layer (user identifiers within the packets) is vacuously easy, ergo irrelevant. In fact, it's not even part of the overall problem.
In certain domains, a user might be a composite of it's source address and additional user information within the packets. Als
Re: (Score:2)
Assigning QOS by user explicitly makes little sense. Not all traffic a person can generate should have the same priority.
An NSA military intelligence officer accessing a satellite mat should have greater priority than his commander surfing the net.
Re: (Score:1)
Assigning QOS by user explicitly makes little sense. Not all traffic a person can generate should have the same priority.
An NSA military intelligence officer accessing a satellite mat should have greater priority than his commander surfing the net.
Well, no one is saying that QOS is to be assigned explicitly by user identity alone every single time and for all domain cases. I'm not sure exactly what on earth makes you think that's the argument, considering the following fragment in the second to the last paragraph in my previous post:
advanced prioritization (which the more I think about it, the more that it might be based on not only on identity and data type, but permissions)
Even w/o considering permissions, it's a fact that QOS is very likely a function of identity/role and data type.
So let me know what part of my previous posts sounded like an argument of assigning QOS by user identificat
Re: (Score:3, Interesting)
$11 million doesn't seem like that much... (Score:2)
It about... (Score:1, Funny)
It's about f-ning time.
$11M v $42M, before anyone asks... (Score:1)
From the article:
"This one-year contract includes two, one-year options, which, if awarded, would bring the cumulative value of the contract to about $42 million, BBN stated. "
Re: (Score:2, Funny)
$42M? Does it include searching for the answer to life, the universe and everything?
Re:$11M v $42M, before anyone asks... (Score:5, Informative)
This has a lot of complicated requirements. If you scan through the pdf "DARPA's Military Networking Protocol" link in the article I don't see how this doesn't extend well beyond 3 years and $42 million. E.G. "As deliverables, performers must provide protocol implementations that replace or modify both the Transmission Control Protocol (TCP) and the User Datagram Protocol (UDP) for the user level devices and the Network Controllers."
Throw in the pace of defense companies move and it would be a miracle.
Re: (Score:1)
This is DARPA... It's been a LONG time since they've ever actually paid for development through final production.
Their strategy is to pay for initial technology development and proofs of concept, and then encourage (and probably help) to find a government customer who actually wants a product based on those technologies.
So the timeframe for a real product winds up being much longer. And the final bill to the taxpayers, much higher.
Re: (Score:2)
Maybe they can get these guys [www.cfos.de] to make it. They seem to have experience with all that protocol/prioritization crap.
The Next Internet? (Score:2, Insightful)
Interesting, could be a precursor to the next evolution of the Internet.
I don't know how well people would like QOS determination on users though, but I see the appeal to the government(s).
Re: (Score:2)
the traffic prioritization thing seems ripe for abuse
Re: (Score:3, Insightful)
Yes. For military use it make sense.
IF we could get enforced guidelines with court protection, it would be fine. Of course we won't get that until after years of abuse from verious corporate entities.
Re: (Score:2, Funny)
Yes. In the future you'll only get Gross Neutrality.
Re: (Score:3, Funny)
Oh! You mean like marking P2P as "low priority" during peak usage hours...oh wait...
Re: (Score:3, Insightful)
Re: (Score:2)
lower level soldiers can get there net access cut back or removed when needed to provide command guaranteed access to the network when they need it.
Ding Ding Ding. This is taking what we already have in the DSN military phone system and applying it to the tactical and strategic networks. This is more of an issue on the tactical Joint Network Node (JNN) networks as the military continues to expand with the entire BCS suite. CPOF alone demands priority networks to work well between sites. As everything gets digitized the network is getting saturated and it cant grow like this without some sort of traffic control. A JNN can take a good amount of load but
Comment removed (Score:4, Insightful)
Re:Opportunity! (Score:5, Funny)
Re: (Score:2)
Re: (Score:2)
That thing was better than I expected. But aren't all politicians nowadays headless servers, that just blindly follow orders from their lobby masters?
Re: (Score:2)
"if the US Gov gets any more czars"
Fuck the Czars, worry about the Rasputins!
Re: (Score:1, Interesting)
so I'm going to dust off my old BBS software and install another landline.
Better yet, get your Amateur Radio [arrl.org] license and practice using Packet Radio [wikipedia.org] and PSK31 [wikipedia.org].
Re: (Score:2)
Re: (Score:2, Funny)
Just upload that spam to missiles and deliver it physically. You surely will get highest priority that way.
is it ipv6? (Score:4, Insightful)
if it isn't, an awesome example of government stupidity, since just as this thing gets off the ground, ipv6 will probably finally take over
it it is ipv6, look for ipv6 to be mandated on the industry
Re: (Score:2)
DoD is already in the process of transitioning to IPv6.
Re: (Score:1)
Looking at the countries where the government doesn't have an effective monopoly on armed force, I really prefer the monopoly in that case. At least as long as it is democratically controlled.
Re: (Score:2, Troll)
Looking at the countries where the government doesn't have an effective monopoly on armed force, I really prefer the monopoly in that case. At least as long as it is democratically controlled.
So 51% of the population can put the other 49% into camps?
The 2nd amendment was supposed to ensure that the government didn't have a monopoly on armed force.
And now.. (Score:5, Insightful)
The corporate think-tanks that envisioned the internet have known for a long time they had unwittingly created a network without strong authentication. This means anyone can jack-in anonymously and spread whatever socially dissident or commie/terrorist agenda they want. So in the interest of controlling our minds and the accessibility of information they are now attempting to re-implement the internet and in doing so shape traffic along arbitrary guidelines which of course will be entirely influenced by corporate profiteering.
I know that this project is only for military use, but it is only a matter of time before corporations are lured in by the promise of an unprecedented amount of power/control/oversight on their networks.
Re: (Score:1)
Re: (Score:2)
Traffic shaping is not necessarily opposed to net neutrality. I see nothing wrong with prioritizing traffic based on how much a customer paid, or how much bandwidth they've used recently, for instance. An ISP account should come with X gigabytes/month of "first class" service, where you get to decide what to send/receive first class, and the rest is bulk. I
Re: (Score:2, Insightful)
If we are to have telemedicine, displacement of copper telephone infrastructure, etc., then we do need guaranteed levels of service for some things.
Traffic shaping is not necessarily opposed to net neutrality. I see nothing wrong with prioritizing traffic based on how much a customer paid, or how much bandwidth they've used recently, for instance. An ISP account should come with X gigabytes/month of "first class" service, where you get to decide what to send/receive first class, and the rest is bulk. I have wasted too much time kluging with LARTC [lartc.org]. Traffic prioritization needs to be end-to-end, not just at the network layer of one end.
You're right. This is just another example of how the technology itself is not malevolent but rather how its employed.
That being said, I can assure you that simply by reviewing the level of divisive manipulation of traditional corporate media, while technological advancement have historically entailed a net benefit for purveyors of truth(thinking here, printing press, telephones, etc), it also has enabled morally bereft institutions to expand and refine their influence on the hearts and minds of the ma
Sounds like a workaround (Score:4, Insightful)
$42M? (Score:2)
That sounds like enough to pay for the answer to the ultimate question of life, the universe and everything!
Isn't "Digital Sentience" a prereq? (Score:2, Interesting)
Iran (Score:2)
This is about censorship, plain and simple.
OK, guys... (Score:5, Informative)
The DoD is big into what they're calling "Network-Centric Warfare". US doctrine relies heavily on information dispersal and access.
This is (currently) an effort to make sure the right info gets into the right hands on the battlefield.
DARPA Error 1205 (Score:1)
This slashdot post was deadlocked on resources with another higher priority slashdot post and has been chosen as the deadlock victim
Terminated (Score:1)
Re: (Score:1)
Does it take that much (Score:2)
to run an On/Off button to the Oval Office?
This way he can take down the Internet to save us all.
Change YOU can believe in.....
Those 1985 .com names are getting snapped up (Score:1)
Who cares how much they spend (Score:3, Insightful)
Shiny tool (Score:2, Insightful)
the problem with this will be that any idiot general or SES exec (as in idiot in terms of comms/technical issues) will be able to order his "comm guys" to make "priority one" all traffic having anything to do with him and/or his cronies. the sycophantic bureaucrats hanging onto this general's/SES's coattails and their hours of grainy video-laden powerpoint slides about the battalion/unit/agency bake sale will crash base networks all over the world. packets carrying beat-the-dead-horse PowerPoint slides wi
32 levels of traffic prioritization (Score:1)
32 levels of traffic prioritization? That sounds like prioritization by bitshift.
I call dibs on Priority 0!
Already been done (Score:3, Funny)
I hope the military enjoys it more than the average peer-to-peer user on cable.
Re: (Score:1)
As any good progammer knows... (Score:3, Interesting)
and will include support for features like 32 levels of network traffic prioritization
...a fixed number of levels means a badly designed program. Or else it would not put any limitations on the number of levels.
Why not just make it go trough the rules recursively like all cascading rule parser? You could even put a configurable limit on it, so it does not crash when coming in contact with infinite levels of rules.
Re: (Score:1)
This is not always true. Consider a real time system. If you want to guarantee response times with bounded computational power, you cannot handle an arbitrary number of priorities. Instead, the number might depend on the available timespan and the worst case execution time of your computation. If
Old Technology (Score:2, Interesting)
> The advanced network technology ... being developed by ... DARPA .. will include
> support for features like 32 levels of network traffic prioritization that will
> let data with a higher priority will be handled more expeditiously than traffic
> with a lower priority
Hahahahahahahahahahaaaaaaa .... "advanced technology" ?
We were doing this in 1980 with the ICL VME mainframe operating system using their proprietary comms protocol "ICLC03", which prioritised traffic according to which of 6 di
Money well spent? (Score:1)