Wireless Networks That Build Themselves

ScienceDaily has an interesting article that looks at ad-hoc wireless networks and how they might be even more useful on a large scale. The RUNES project is featured as an example of software projects that might be able to make mobile devices that form self-organizing wireless networks to help promote this goal. "RUNES set out to create middleware: software that bridges the gap between the operating systems used by the mobile sensor nodes, and high-level applications that make use of data from the sensors. RUNES middleware is modular and flexible, allowing programmers to create applications without having to know much about the detailed working of the network devices supplying the data. This also makes it easy to incorporate new kinds of mobile device, and to re-use applications."

Comment removed

[account_deleted]

Comment removed based on user account deletion

Re:Responsibility

[ColdWetDog]

What happens when the mobile devices get together and watch "Terminator"?

Re:

[Brian Gordon]

What happens when they start getting included in major routing tables and when a backbone goes down, a lot of data's going to start moving through people's devices.. just think of slashdot, no secure login..

Re:

[SatanicPuppy]

Encryption still works the same over this sort of network...Doesn't matter if someone in the middle reads your public key, the communication is still encrypted.

Still, I was thinking encryption would be necessary for basic privacy...Something like Tor, where you don't know who is requesting what data. Otherwise it'd be too easy to figure out who was downloading what porn in your neighborhood.

Re:

[Asm-Coder]

Yes encryption will still work, but some websites (read Slashdot) do not have secure login facilities available.

Re:

[DragonWriter]

What happens when they start getting included in major routing tables and when a backbone goes down, a lot of data's going to start moving through people's devices.. just think of slashdot, no secure login..

What's your point? The regular backbone is operated by the telecom industry, which has demonstrated willingness to open it up to the government even when that is in direct violation of existing law. Your unencrypted content isn't safe no matter who owns the network it travels over (unless you control bot

Re:

[Fred_A]

What's your point? The regular backbone is operated by the telecom industry, which has demonstrated willingness to open it up to the government even when that is in direct violation of existing law. Your unencrypted content isn't safe no matter who owns the network it travels over (unless you control both endpoints and all the systems in between.)

That's the "environmental monitoring" feature cited in the article I presume.

Re:

[guardiangod]

You will get many badly imitated "Hasta la vista" in pseudo-Assuie accent.

Re:

[urcreepyneighbor]

What happens if your mobile device forms a node over which someone else gets child porn?

You're fucked?

Re:Responsibility

[iknownuttin]

What happens if your mobile device forms a node over which someone else gets child porn?

FTFA:Applications include emergency management, security, helping vulnerable people to live independently, traffic control, warehouse management, and environmental monitoring.

I really don't see this protocol, at first anyway, being used for consumer devices. I'm sure someone will find an application for it, but I don't see the need in the near term. And, I would assume, there would have to be some sort of identifier of the sender and ultimate receiver like TCP/IP has in its protocol.

Re:Responsibility

[SatanicPuppy]

Nothing. Common carrier.

The day when this becomes real will be the day that traditional ISPs die. The only way to reliably monitor the traffic will be some kind of "seeding" where the monitor-er will put out relays that monitor the traffic that passes across them.

I think this is really possible in the long run, but in the short run I don't think most things have sufficient computing or broadcast power to make it a reality. Cool that they're working on it though.

Re:

[sltd]

I don't know that it would completely kill ISP's. There are some people who live in really remote places, or at least too far for this kind of a network to be worthwhile. Some areas still don't even get broadband access, because there aren't enough computers for it to be viable. Would the people in these areas just be cut off?

Re:

[SatanicPuppy]

Yep.

Heh. No I have no idea; I'd imagine that repeaters would be set up in a lot of places, just to lighten the load on consumer devices. How they'd be funded, I have no idea. It'll all depend on the eventual power consumption, and what parts of the spectrum are used.

Re:

[SatanicPuppy]

I'm in favor of that for hard lines, so obviously I think it's great for wireless too. At this point having telecoms owning all of our information architecture is getting pretty scary.

Re:

[j1m+5n0w]

Ad-hoc routing doesn't get around the need for ISPs; someone has to have a gateway to the Internet. (And as long as ad-hoc routing doesn't scale up to networks of billions of nodes rather than hundreds or maybe thousands, we're going to have to peer with the Internet somehow.) ISPs may have fewer customers, since a lot of them will be leeching off of their neighbors, but most people who can afford it are probably going to go ahead and pay for their access, since it's faster (multihop wireless networks aren

Re:

[westlake]

Nothing. Common carrier.

The common carrier is an organization or enterprise that provides messaging services to the general public.

It is not about the tech.

It is not a defense against trade in pornography that you can claim by right.

You are a common carrier if - and only if - you meet the statutory definition and requirements and your conduct remains within the law.

The geek is far, far, too enamored with the idea that technical competence - technical innovation - puts him out of the reach of the law.

Re:

[navtal]

1.Create a wireless P2P network. Every device is a wireless router.

2.Give it a hardwired bandwidth so it dose not overwhelm the users bandwidth.

3.Give it two completely different network cards. One for the users needs and one for the wireless routers roll to secure the users or the network from being attacked or corrupted by each other.

4.Create appropriate legislation protecting information carriers from the actions of the distributors and producers of illegal content.

5.????

6.Profit

Re:

[account_deleted]

Comment removed based on user account deletion

Re:

[bemo56]

Wasn't the OLPC supposed to come with a wireless network that achieved this already? http://en.wikipedia.org/wiki/OLPC_XO-1#Wireless_mesh_networking [wikipedia.org]

Such a thing could easily be implemented by a piece of software running on everyones laptop, think of it like P2P only with FREE INTERNET!!!!

Hmmn, Seems I have found myself a coding project.

P.s. how do i start a new thread instead of always replying to first? :S n00b :P

Re:

[Fred_A]

Wasn't the OLPC supposed to come with a wireless network that achieved this already? http://en.wikipedia.org/wiki/OLPC_XO-1#Wireless_mesh_networking [wikipedia.org]

Such a thing could easily be implemented by a piece of software running on everyones laptop, think of it like P2P only with FREE INTERNET!!!!

Hmmn, Seems I have found myself a coding project.

/quote>
Since it has already been coded and is GPL, why do you want to recode it ??

Re:

[bemo56]

Since it has already been coded and is GPL, why do you want to recode it ??

Well that wouldn't be as fun, although i have to admit i overlooked that :P

I'll bet....

[UncleTogie]

...that malware writers will LOVE this. Free propagation, just add mesh!

Re:

[TheSpoom]

Nah. I bet this could be designed in such a way that NAT still exists behind the router; the routed traffic can't actually get into the internal network of the users' PCs and devices unless they've specifically port-forwarded something in the router. Most malware worms spread by exploiting holes in Windows' WINS / file and printer sharing / etc. and if these ports are closed, there's no way they can get in (without breaking the router).

Re:

[UncleTogie]

I bet this could be designed in such a way that NAT still exists behind the router; the routed traffic can't actually get into the internal network of the users' PCs and devices unless they've specifically port-forwarded something in the router.

The idea of multiple propagation routes for malware of all types {and in which category I also lump viruses...} has already been done... [f-secure.com]

Re:

[SatanicPuppy]

It'd be no different than IPv6, with every node available.

No mistake, in the long run we're going to have to have significant advances in security, because we won't be able to segregate every vulnerable machine behind a big security infrastructure.

in the words of bender..

[pak9rabid]

neat

Recipe for Cash

[uberhobo_one]

Step 1: Set up a node(s) at the very edge of a mesh network. Step 2: Install software to execute a man in the middle attack. Step 3: Wait for someone to connect to you alone. Step 4: Wait for that someone to connect to their bank. Step 5: Drain their account. It'll take some clever protocols to prevent abuse if this ever gets used as a standard consumer network protocol, but it should do wonders for emergency services.

Re:

[MightyYar]

Wouldn't the bank still have their signed key? How is this different than the situation now, where you could sit at a coffee shop and impersonate a free router?

Though I guess people might just accept any old key that gets thrown their way, ignoring the warning that comes up...

Re:Recipe for Cash

[SatanicPuppy]

Just because someone routes traffic through your node, doesn't mean you can read it. A man in the middle attack across encryption requires that people accept unsecured certificates, which happens often enough, but it's not a slam dunk by any stretch. Theoretically they could try to screw with the PKI, but that would involve breaking their keys, and if you could do that, then stealing someone's bank info would be trivial.

The great thing about public key crypto is that the key that is visible is meant to be.

Re:

[mrogers]

If only SSL had been designed to make man-in-the-middle attacks impossible. Oh wait, it was! Your browser contains root certificates that are used to verify that your bank's certificate hasn't been replaced or modified by an attacker. MITM attack against SSH? Maybe, if you don't check the key fingerprint (and I doubt anyone does). MITM attack against SSL? No chance, unless the server has a self-signed certificate, something no bank would consider.

Please consider...

[jd]

...that a few years back, it was widely reported that someone was able to get Verisign to provide them with Microsoft's server-side certificates. I don't think it was ever really said how long the rogue copies were in the wild before action was taken. I can't see banks ever admitting to their security being compromised in such a way.

Having said that, phishing is still described as highly successful, social engineering is usually described as highly effective, and there are probably a few places where ward

Re:

[mrogers]

I have no idea why you got modded down for that, it's a good point. The same applied to self-signed certificates if you save them the first time.

ISP-less internet topology

[Prysorra]

The internet was designed to route around bottlenecks and network damage. ISP control is a type of bottleneck regardless of the amount of bandwidth.

Thus a natural progression to further decentralization is exactly what is happening. Expect to see ISP trying to pressure legislators to ban this kind of technology, and spreading FUD about.

Re:

[irregular_hero]

The Internet has proven to be more than slightly "put off" by a well placed boat anchor, I think.

A more correct statement would be: Due to the extremely high traffic loads some core links carry and the inability for backups to primary network paths to keep up with the growth rate of network traffic, the internet was designed to limp around shouting "ow, ow, ow!" around bottlenecks and network damage while IT staff groan about another sleepless night of babysitting outsourcing engagements whose bandwidth is

Re:

[geekoid]

"The Internet has proven to be more than slightly "put off" by a well placed boat anchor, I think."
oh? everyone seems to still have access. Which incident are you referring to the brought down the internet?

RUNES Homepage

[sconeu]

Runes Homepage [ist-runes.org] for those who want more depth.

What a Wireless Sensor Node is:

[corsec67]

A wireless sensor node like the Tmote Sky [sentilla.com](pdf) is a very tiny embedded computer that runs on 2 AA batteries, and is usually the size of the back of the 2 AA battery holder. They have a radio on it, but the radio isn't compatible with 802.11b instead compatible with 802.15.4 [wikipedia.org], and is limited to about 256kbps. The Tmote Sky has a 8MHz 8-bit processor (the Atmega 128), 10KiB of ram, 1024KiB of flash, with a few A-D inputs and some digital outputs. It isn't exactly very fast, nor does it have a bunch of ram.

It is designed for a distributed sensor platform, and not doing a lot of computation.

A picture of one is here [flickr.com], connected to a 14-foot USB cable.

Re:

[corsec67]

Replying to myself to add some more information about the Tmote Sky and similar wireless sensor nodes:

The main problem with these sensor nodes right now is mainly that they are just way too expensive for what they are, at $140 each. Since any application of them is in large numbers to get around the 10-meter radio range, it gets very expensive quickly to do anything really useful with them. That is $140 for a device that doesn't have a screen, case, keyboard or external antenna.

The goal is to get the cost u

Whats in a name...

[Darth Eggbert]

I came up with a concept one night for a Fractaly Organized Nearby Transient Area Interface NEtwork (FONTAINE) in which each device had 3 Transceivers which would connect to a different device on the network, which in turn connected to 3 other devices etc. it would be infinitely scaleable and each device would carry a map of all the connections in the network. Once logged into the network your device would constantly search out the strongest signals and update transceivers one at a time. If one node gets ov

Re:

[irregular_hero]

"Infinitely scalable" AND "carry a map of all the connections in the network"? As # nodes approaches infinity, the number of map entries would approach, what... infinity ^ 3? And the amount of memory needed to store a, say, 128 byte record of each routepath would require memory on the order of infinity ^ (3 * 128)?

I do not think that word means what you think it means.

Not so simple

[stevedcc]

The article talks about everything from motes to handhelds, all on the same network. I work for a company that has a low-bandwidth low-power sensor node product, selling software to hardware makers, and hardware for prototyping purposes. The requirements vary so much from sensor-only devices to handhelds, that any product catering for both would be inherently compromised. Does your handheld want to work with a network that has a total bandwidth like modems from 20 years ago, shared between all the nodes? Is it really concerned with keep power emissions so low that it can stay on that network for 10 years, powered by batteries? how about a sensor attached to your radiator?

Techies tend to think about what CAN be done with a certain technology, but sometimes we try and generalise too far

Zigbee

[us7892]

Sounds like open standard *Zigbee* http://zigbee.org/ [zigbee.org] networks. Been hearing about Ember http://ember.com/ [ember.com] chipsets and self-healing, self-discovery wireless mesh networks for a few years now. Pretty quiet as of late.

Some links

[klapaucjusz]

Some links, if you want to learn more, including some astroturfing:

DSDV [wikipedia.org]
OLSR [olsr.org]
AODV [wikipedia.org]
Babel [jussieu.fr].

Terminator has to kill this concept

[LM741N]

If you have watched the Terminator movies you know that it was self organizing and linked computers that became sentient (sort of) and decided to kill off the human race. Truth is stranger than fiction. Nowadays I guess they can just link up all the voting machines and pick a computer as a write in President.

Peer to peer mesh networks are a dead end

[flip314]

Although TFA doesn't say whether these are peer-to-peer networks, I suspect they are. The best theoretical aggregate bandwidth you can get out of a set of mesh-connected peers scales as sqrt(n) in the number of nodes, which isn't that great. In most practical mesh systems, the total bandwidth goes DOWN as you add peers. Until somebody figures out a better way to organize ad-hoc mesh networks, they really won't be all that useful for most applications.

nonzero bandwidth better than zero bandwidth

[j1m+5n0w]

Any bandwidth greater than zero over a multi-hop link is an improvement over the status quo, in which that sort of thing typically just isn't supported. Ad-hoc networks are slow (I'm not sure where the sqrt(n) comes from), but they are very practical in situations where a star topology isn't practical, and you want something that just works, right now, without a lot of manual configuration.