NSF Tags $30M For Game-Changing Internet Research

coondoggie writes "So you want to build a better Internet? The National Science Foundation today said it would spread $30 million over 2-4 projects that radically transform the Internet 'through new security, reliability and collaborative applications. The NSF said its Future Internet Architectures (FIA) program wants: "Technological innovations and the requirements of emerging and yet to be discovered applications, the Internet of the future is likely to be different from that of today. Proposals should not focus on making the existing Internet better through incremental changes, but rather should focus on designing comprehensive architectures that can meet the challenges and opportunities of the 21st century."'"

Comments

Likely to be different?

[Chris Burke]

So, the internet of the future isn't going to be a general-purpose protocol-agnostic world-wide data network for sharing and communication of information?

Uh, can I opt-out of the future?

Re:Likely to be different?

[FooAtWFU]

I predict the next big thing for the Internet will need to wait until Google rolls out its version of a communications security infrastructure, issuing people certificates (why not? they know enough about you already) and helping them with public-key cryptography, ultimately leading to an email system free of spam.

Some decade.

Re:Likely to be different?

[Chris Burke]

and I'm not talkin about what oysters sometimes produce.

You're not talking about sperm? Okay now I'm just confused.

Re:Likely to be different?

[icebike]

If only the future had opted into the past.

Quote from TFA:

From the Network World article: The NSF says it won't make the same mistake today as was made when the Internet was invented, with security bolted on to the Internet architecture after-the-fact instead of being designed in from the beginning.

"We are not going to fund any proposals that don't have security expertise on their teams because we think security is so important," says Darleen Fisher, program director

And this really is the crux of the problem isn't it?

Rampant SPAM (95% of all email), deep packet inspection, attacks, bot nets, the list goes on. Almost all the abuses we suffer daily on the internet are due to the security-as-an-afterthought model.

There will be those (there always are) who insist that this is nothing more than a government take over and the installation ob yet more back doors. There is nothing that can be done to appease that viewpoint, even open standards and open source will not suffice.

But I am not prepared to believe we can not improve upon what was done 40 years ago given the number of minds and the level of technology we have to apply to the problem today.

We defend the status quo because we know it, not because it is optimal, not because it is even close to being fully functional, and certainly not because it is fair.

Deal with political problems in the political arena. But in the mean time, lets fix our tools.

Re:Likely to be different?

[Chris Burke]

Rampant SPAM (95% of all email), deep packet inspection, attacks, bot nets, the list goes on. Almost all the abuses we suffer daily on the internet are due to the security-as-an-afterthought model.

Not really.

Bot nets exist because you can never stop people from installing software no matter how scary your warning dialogues about untrusted sources are (and in fact throwing up too many is counter-productive).

Spam and DOS attacks are because you can't prevent the bot nets.

Most of the real security problems are at the OS/application level. Not the underlying internet.

Re:Likely to be different?

[misnohmer]

You may like this [http://www.nebunet.com] - social networking of any IP connected devices, not just people. The idea is to turn the internet into many independent secure networks as easy to use as your favorite social networking site. It's not something google would like to see - self organizing internet based on context - but most people would. What do you think?

Re:Likely to be different?

[Chris Burke]

No, it's because there aren't many security problems to solve at the IP layer or below.

You can't stop botnets or spam by putting security into the internet itself. Not without breaking what the internet *is*.

Re:Likely to be different?

[Chris Burke]

Who says a new design has to use IP?

So... you're planning on introducing a bunch of security problems below the transport layer?

You'll still have to solve all the problems again at the application layer!

Remember, at the time it was designed, there was no "is".

Yeah instead there was a "designed to be", and it was designed to be what I described in my first post. You can break that if you want. I like it.

Re:Likely to be different?

[DragonWriter]

But I am not prepared to believe we can not improve upon what was done 40 years ago given the number of minds and the level of technology we have to apply to the problem today.

We can, quite easily (on the technical front), but it doesn't take any stunning new transformative technology, just the kind of incrementalism that the effort here disdains. Its not like the problems of SPAM and other similar problems haven't already spawned technologies designed from the ground up as complete "super-replacements" (that is, replacements with broader general applicability than the replaced system) that are also designed to avoid the problems with the replaced systems. For email and the problem of SPAM, AMQP (a generalized messaging protocol which subsumes, but goes far beyond, the function of email) is designed from the ground up to avoid the possibility of recipients being spammed.

The problems with replacing existing technologies with more secure ones is more of a social problem than a technical one. Putting money into technical research that specifically requires that it go only into things that are radically different than what exists now -- and thus a bigger social problem to get people to transition to -- don't help at all.

Re:Likely to be different?

[raddan]

Actually, rethinking global addressing schemes is on the table for many next-gen Internet projects I've spoken to researchers about. The reason is that router-table growth is not adequately handled in IPv6, nor is the meaning of an IP address very clear in the current Internet. These are major issues. Have a look at Jerome Saltzer's work on naming and addressing. If you want the short version, have a look here.

SPDY

[Enderandrew]

http://dev.chromium.org/spdy/spdy-whitepaper

Do I get $30 million for finding that for him?

Step 1:

[swanzilla]

Abolish Flash, immediately.

I can solve this easy

[antifoidulus]

through new security, reliability and collaborative applications.

No need to create new tech to do that, I can increase the security, reliability, and the collaborative potential of the internet easily, just get rid of Windows. There, can I have my $30 mil now?

Re:I can solve this easy

[antifoidulus]

When the Chinese hackers decided to go after Google, which machines did they go after, the Linux servers or the Microsoft Windows clients? Answer, despite the fact that the data they were after lives on the servers, they went after the clients because Microsoft "security" is a joke and serious, easy to exploit holes go unpatched for months on end from Redmond. Not to mention the sheer amount of shit they REQUIRE you to be an admin for, the total lack of opacity in their processes etc. If Microsoft disappeared tomorrow, there still would be security exploits, but significantly less than there are now.

Not to mention the numbers speak for themselves, despite having over 90% of the pc market share, Microsoft has less than half that and that share is continuing to decline. Why is that? Because cracking Windows is pretty trivial, and if a company has important data they want to protect, they sure as hell aren't going to go Windows.

Microsoft has never paid anything but lip service to security, and I suspect they never will. Oh well, the sooner the world is rid of Windows Server, the better.

Re:I can solve this easy

[causality]

I don't know if you are an arrogant Mac user or a Pompous Linux Guru, but you have to realize that the vulnerabilties in Windows do not make the FUNDAMENTAL vulnerabilities in other systems go away.

If Microsoft folded up shop tomorrow and the only Machine you could get at a big store was a Mac, one of two things would happen. Either A) More and more viruses would pop up for Macintoshes. And yes, there are some, so don't try and deny that. Or B) Macs, being locked into a very specific hardware set would have to adopt a more open policy (opening more holes) or It would cause some serious stagnation in the producers of other computer parts - completely ruining all competition and slowing all progress.

And if everyone were using Linux, it would be just the same as before. Everyone would be Sudo'ing this and that and hackers will exploit any setup the user uses to make their PC Easier.

You need someone like Microsoft to be the scapegoat for the idiot masses so that more secure systems can even exist.

Microsoft is just catering to a need. The "need" is that people want to use technologies and networks without understanding what they are using or at least learning about their correct use. So long as people think this is a great idea and refuse to invest a little time learning about the tools they use every day, the security situation is not going to improve. I'm actually fine with this; people who fall for phishing attempts and the like are merely getting out of the system what they were willing to put into it. It concerns me that this is not a technological problem but technical solutions are being proposed for it. Those can only have the effect of restricting the free and open network that is available today for anyone who wants to learn how to use it.

This is nice

[dedazo]

But honestly, with the US so far behind other industrialized nations in broadband quality and penetration, shouldn't this be promoted by Japan or South Korea? Who cares about the super duper better intertubes if you're still stuck at the 1.2mbps downstream dictated by the local suckage cable mini-monopoly?

I'm all for this type of thing, I really am. But fix the basement before you go adding a new chimney.

Re:Time to disolve NSF?

[goose-incarnated]

There is much better use for 30M such as spending it on education, which is broken rather than Internet which isn't not so broken.

Yup ... you're seriously making a great case there, trust me on this ;-)

I'll take cash or check...

[cosm]

Security:
Fourier Transform FT( Internet ) - Security through obscurity, it won't make any sense!

Reliability:
Mobius Transform MT( Internet) - You always end up where you start, SynAckishly

Collaboration:
Wavelet Transform WT ( Internet) - Make it a design ideology, Google's got it ;)

Best to keep doing patches

[Darkness404]

Its a lot better for the world as a whole if we keep doing small improvements to the internet rather than a total overhaul. For one, it will create a -huge- amount of waste in a short period of time, for another, it will not be entirely global, corporations, governments, etc will aim to reduce global communication, global trade and such. If we do create a "new internet" it should be decentralized as much as possible, nearly untraceable and fully global (no Geolocation-IP address based discrimination), however, governments do not like us to exercise any freedoms they have on paper and corporations want to maximize profits, so this will never happen.

Hey, I'll take a couple hundred grand for this

[goose-incarnated]

I doubt that this is open to non-Americans, so I'll just post my idea here instead:

Make every endpoint (home 'puter) have no less than two different ISP connections. Then every home computer can also be a router. This does mean that every single packet has to be encrypted (a solved problem, methinks), and that every single endpoint is properly uniquely identified.

Advantages are numerous - encryption is required for it to work at all, consumers have redundancy (not only for their own net connection, but throughout the entire path as well), ISP's don't have to provide $X Mb/s connection, they can provide $X/2 Mb/s and the computer can load-balance while routing. Last advantage is that torrent-like downloads can take place without the need for special p2p software.

Disadvantages do, of course, include the fact that every consumer doubles their internet bill and that a govt is unlikely to fund a global TOR rollout :-)

A series of tubes!

[gestalt_n_pepper]

Oh wait, somebody already took that one.

More hot air.

[Cryacin]

It is very unlikely that there will be a radical change in the Internet. Too many businesses, governments and people rely on current standards, that are going to be disruptive and expensive to change.

Don't believe me? Look at your power socket. Not many countries change their standards, and it requires quite a bit of expense to make it happen. Hence why manufacturers and consumers wind up looking stupid when bringing an American appliance to a European power socket. *SCHMOKING!!!*

Adoption

[cosm]

Wishful thinking. What makes them believe anybody will adopt? The general theme I gather from the Slashdot community is that the preexisting design aesthetic (if you can even call it that) for the internet is actually pretty solid, its just the implementation that people & organizations botch. The IPv6 bandwagon isn't about to collapse from all its passengers now, is it?

The folks who generally engineered the internet had decent enough foresight from a technical standpoint. It is the BIG Telco's and all their 'peering', 'filtering', 'throttling', and combined unwillingness to invest in new infrastructure that puts the choke hold on our tubes (pun intended). Do you expect the major Tier 1's to drop billions of $$$ to adopt, 'cuz I sure as hell don't.

why spend all that cash?

[mt1955]

Wouldn't it be cheaper just to call Al Gore?

Re:Time to disolve NSF?

[Truth is life]

There is much better use for 30M such as spending it on education, which is broken rather than Internet which isn't not so broken.

That's not the point of the NSF. Besides, as this link http://nsf.gov/pubs/2010/nsf10001/toc.jsp to their FY 2009 report shows, they already spend almost a billion dollars a year on education. Or over 30 times the value of this award. I really don't think you can claim that canceling this award and giving the money to the DoEdu (or even shifting it to the education side of NSF) would be better value for the money.

Whose security are we talking about?

[CopaceticOpus]

Increased security, built into the fabric of the internet, sounds like a goal everyone can support. However, to build security into the network, you must necessarily build in stronger methods of identifying the users of the system. This will make anonymity much more difficult, and will greatly increase the government's ability to track the online activities of individuals.

There are some situations where that power would be used for good, but do we really want to allow the government more power and more ability to monitor the population? I am sure that they are drooling over the possibility. The recent abuses of the FBI should give everyone a fair idea of how responsibly this power would be used.

I'm not sure what a "game-changing" technology would look like, anyhow. The internet is fundamentally about shuffling bits of data between endpoints. That much is not going to change, and the rest is just implementation. What are we going to try, sending twos?

Re:Time to disolve NSF?

[TimHunter]

Don't waste it on education. $30M is much better spent fighting hunger. And working for world peace. Spend the $30M fighting hunger and working for world peace. And manned space exploration. Spend the $30M fighting hunger, working for world peace, and manned space exploration.

I'll come in again.