Slashdot Mirror
OLPC Mesh Networking Tester Explains How It Works
An anonymous reader writes "James Cameron is an engineer working on the OLPC project, specifically testing the wireless network capabilities of the OLPC XO laptop. Cameron lives in a small town called Tooraweenah in a remote region of the Australian outback. There is little noise in the spectrum in the area, so it's perfect for testing the wireless networking capabilities of the XO as it mirrors the kind of rural, spacious environment the XO is intended to be deployed in. Cameron breaks down exactly how the OLPC XO's mesh networking works, including the cheap US$35 solar powered mesh nodes that can be mounted on top of a tree to further the network's reach. Testing in the Australian outback, Cameron discovered that the range of the XO could go up to 1.6km 'quite easily' at 1.5m above ground. 'Assuming a range of 1.6km holds true, (the mathematical formula for area of a circle) Pi R squared tells us one well placed mesh node will cover up to eight square kilometers.' The article also includes numerous pictures of the mesh nodes and testing of the XO."
I found this review to be logical and informative and therefore boring. So I decided to let a few key individuals weigh in on this and tell us exactly how this mesh network works.
... DRIVE over, no, it's a junction for that internet and also your own personal internets. See, just the other day, I got on this here thing and it sent out what I call a "searcher tube" looking for other laptops ... up to 1.5 km away. That's right, once a searcher tube finds another larptop or 'junction tube' then it connects to that and the series of tubes continues to grow. Unfortunately, this series of searcher tubes and junction tubes makes it highly probable that my messages get backed up in those tubes. Therefore, this will provide an own personal internet for poor children in countries we either need to invade or ignore but it will not, however, suffice for bridging islands in Alaska."
... and it was good! Upon God's recognition of my authentication of that which you call "the login screen" a hand descended from the heavens. This hand stretched 3 km end to end and it was then clear to me then that this hand was intended not for me ... not for the rich ... not for the privileged but for the poor and pathetic chillun' of the world that need God's help. Now I have a method by which to contact them and teach them about God and ask them for weekly tithes! And when I realized that a TCP/IP connection had been made, I fell down on my knees and prayed to the Lord God for He is Good and Holy and brings life to these innate objects you call the XO laptop, Amen."
... so anyway the laptop like wanted to be alive so I hit the power switch, man, and then it was like *wham* *whiz* *whazzle* and suddenly like the thing could 'talk' to other things like 1.5 km away and I wasn't sure if it was the uppers I had just taken or if I really was connectin' to another machine or network that far away.
... no steal music at their leisure from a vast expanse of 1.5 km. They may already be using these laptops to steal your diamonds and moneys. Putting an XO laptop into your child's hands is only a gateway to crime and an early death. It's cheap price makes it easy to manufacture and distribute, we are here to warn you about the looming threat of a solid network of these devices stretching out across our free country. Thousands of children in other countries have received them already and their quality of life has plummeted. These laptops are sturdy and brightly colored, they are easily identifiable. Find people that have them and report them to the RIAA."
Senator Ted Stevens: "This here larptop isn't an intersection for trucks to just
The Reverend Billy Graham: "And lo, I did with God's good graces ask for power to be restored and replenished throughout the XO's motherboard thereby bringing the only free BIOS to life
Bob Dylan: "Yeah, ok so like, I got this laptop and it was pretty groovy but I had to put my Mac down because it was like I couldn't use two laptops at once
Mitch Bainwol: "We have discovered that a new technology exists that is a threat to your safety & the economy and will further destroy our income in the near future. It allows criminals, drugs users and child molesters to contact each other freely and unmonitored up to 1.5 km away. They can trade
My work here is dung.
James Cameron is working on computer networking technology? I think we see where this is going...
Give a man a fire and he'll be warm for a day. But light a man on fire and he'll be warm for the rest of his life.
Why not chose 0.5-1 metres, as that would be more in line with where the machines will normally operate in hands of kids and on tables.
c++;
Everybody says "Pi R squared", but pi are round. Cake are square.
No folly is more costly than the folly of intolerant idealism. - Winston Churchill
If you haven't made a developer cry, you've wasted a day.
This would be a boon to those of us that are maintaining a community wifi setup. Instead of having to spend a LOT on each node we could easily set up hundreds of these cheap $35.00 repeaters to give users the ability to have wireless throughout the town or city. right now I have dish antennas linking sites and having to buy other gear to get things working for the community Wifi. if I could spend as little as $120.00 per site for a repeater+solar panel+battery and simply get them installed all over the place within range of each other it would be dramatically easier for me. I can put up an unobtrusive box+panel easier than getting permission to put up this box and dish, oh the dish needs to be high up in the air and visible... etc...
They really need to release the whole shebang to the world so that windows drivers can be written to use that mode, linux and OSX drivers would be great too, plus get people making the repeaters better stronger and cheaper.
did I miss the links? do they release all the details of this so It can be implemented commercially?
Do not look at laser with remaining good eye.
In a wireless mesh network, 99.9% of which are implemented using half-duplex transceivers, once past your immediately adjacent nodes, for each additional hop away you halve the bandwidth and double the latency. No way around that, and in a large area mesh (think muni wifi here) you must re-insert an additional backbone feed about every 4 to 8 hops into the mesh either by landline or by point-to-point wireless bridges, else performance across the mesh gets abysmally poor, very quickly.
Dense meshes just don't work very well, they implode upon themselves. Very sparse meshes, such as used in the battlefield by our military, of perhaps in remote areas like the Aussie Outback as mentioned in the FA, are ideal applications of a wireless mesh network, but all the folks who think they can make a successful commercial venture with a wireless mesh in a dense urban or suburban environment are in for a rude awakening if they drink too much of the Koolaid hype that many of the consumer-grade hardware vendors are trying to push.
For those who want to build their own mesh, check out the open-source ROBIN project. They are building a complete plug-and-play mesh networking package. they are even configured to automatically connect to the Open Mesh Dashboard so you can manage your network. Open Mesh will start selling pre-flashed nodes this week at their site.
'Every story, if continued long enough, ends in death.' --Ernest Hemingway
I'm been slightly interested in mesh add hock wireless networking for a while now. I'd like to see some combination of wireless access point in Wifi & WiMax with some sort of Bittorrent management / bridging function. If you could put this in a box like what OpenWRT runs on and sell it for less that 100 it would be great. Living in a college town I'm convinced those kids are sharing petabytes of Porn, Pop & House music, and cheesy serial television shows from Hollywood. Having all that sharing pushed off the wired net I pay my ISP for on to a wireless grey market mesh network would free the wired network up... for my *legitimate* traffic of Blues & Classical music, and David Attenborough documentaries... then perhaps I could finish seeding my fair share of "Life in Cold Blood".
Nothing in the world is more dangerous than sincere ignorance and conscientious stupidity.
But....Pie R Round, not square, so the equation should be:
Area = Pie(Crust + Filling) * Hungry Children + Sticky Keyboards
Area = OLPCs covered with Pie Filling
Area = 0
No. Its just you. The internet decided to pick on you today.
It reminds me a lot of the early Gnutella days when it used broadcast routing.
Bottom line - OLPC network mesh software is pre-alpha.
From the article:
"The school might have a generator or a solar panel, or in one school where we've got laptops deployed now we have two cows who walk around pushing a lever which rotates a generator that powers fifteen laptops for charging"
Shouldn't they be using a couple of penguins instead?
To test in the Australian outback sounds like a test under ideal conditions. No RFI. No natural or man-made obstructions. No problems with climate or weather.
Maintaining "hundreds repeaters" through a Buffalo winter presents a somewhat greater challenge.
What does that have to do with the stated purpose and use of the OLPC mesh networks? It doesn't. Yes, they are a bad idea for urban areas, but people in urban areas are most likely not going to be buying OLPC's.
Mod Parent Offtopic.
A no or low cost Autonomous consumer owned telecommunication infrastructure is what will evolve out of this. No more cable, internet, or cell phone bills.
"an infinite player that has lost his finite mind" ~Infinite Play the Movie (it blends with reality)
1.6**2 * pi is about 6.28, isn't it?
Dense meshes just don't work very well, they implode upon themselves.
I heard the same argument in the early days of the internet. If we let just anyone on the network, requests will slow to a crawl.
Mesh is a fairly new technology, I'm confident the density issues can be mitigated. So, yeah, a dense mesh doesn't work well today. But to me that's not a hugely difficult problem to solve. Latency may be more of a challenge, as you also pointed out. Still, it's Gen I, give them a chance. Reminds me of people harping on Linux two or three years ago.
Telcos and other providers should tremble at the thought of mesh networks. Network technology that doesn't need them to spring up and function, almost anywhere. Self-discovering local phone networks...that'll keep AT&T up nights.
That's our life, the big wheel of shit. - The Fat Man, Blue Tango Salvage
Just because the OLPC is designed to use the entire WiFi band for its mesh network *does not* imply that it's not a mature design, just that it wasn't designed to co-exist with other WiFi networks on the same band.
-- The act of censorship is always worse than whatever is being censored. Always.
It's not the internet, it's teh t00bs.
You might have a small village or education campus implement a community-maintained infrastructure of nodes - but you will not replace mainstream ISP's or Cell providers. That would require everyone making a grand unified switch to mesh networking suddenly at once. Also, who is going to pay for the mesh nodes every 1.5km to cross the oceans in your free-telecommunications-utopia?
If this takes off [at all], at best expect the equivalent of "free wifi!" in a few isolated towns/campus's. This is simply the next evolution of existing 802.X tech.
Trackball users will be first against the wall.
You're not down on mesh networks, your down on urban mesh networks. Courtesy requires that you title your post accordingly. I mean, is there any place less urban than the Outback?
You are new to life in general, aren't ya? Didn't anyone tell you that humanity, as a collective, is a lazy and sheep-like group? The autonomous consumer telco is the urban fantasy that is right up there with growing your own vegetables, weaving/sewing your own clothes, and "living off the grid."
Sure, there are examples of people doing just that, but how many others are there that simply want to live without the "hassle" of being self sufficient.
2 points:
:(
-- the OLPC's were idle for God's sake! They were open and on. Why in that circumstance should they be DoSing the network with neighbour detection code. Sure, when they first open you want to find your neighbours, but once you know about them - shut up! They should only transmit after initialisation when they have something to communicate!
-- there were 4 of them. I will lose my membership for not knowing this, but aren't there 11 channels (not sure how many non-overlapping channels though). How can 4 machines DoS every wifi channel when they aren't even doing anything!
I stand by my point: the network code is pre-alpha and all the smart geeks at the meeting agree. (air-stream.org is a geek-only community wifi project).
ps. the last few days slashdot.org has developed an almost infinite loop every couple of refreshes. The status bar seems to blame it on continually downloading from ad.doubleclick.net. I usually hit refresh to try again, but that doesn't work when you want to post comments
I've seen the same. I wonder if they're doing something with cookies or ip tracking to karamize the anonymous?
Dude. Mesh is not for synchronous communication. Centralized, synchronous services like WWW just aren't going to happen on mesh.
What can happen is something a synchronous like Usenet or E-mail. You could even supplement the existing network with vehicle-mounted hot points. Postal trucks, mobile health clinics, bookmobiles, and other services make the rounds regularly. No reason why they can't spool or relay messages at each stop.
Besides, centralized services like WWW are too easy to censor. Mesh can help drive a new round of freedom of communication, if it can steer clear of proprietary codecs and formats entirely.
Beta is broken and the link to classic doesn't work. Stop wasting our time or there won't be anybody left here.
In the US, the 2.4 GHz ISM band has 11 channels spaced 5 MHz apart. 802.11b and 802.11g require 25 MHz of separation to prevent interferance which limits the non overlapping channels to 1, 6, and 11. 802.11n and many 802.11g systems support double channel widths of 40 MHz which limits the 2.4 GHz ISM band to just one non overlapping channel.
The 5.0 GHz band used for 802.11a and for some 802.11n radios has 19 20 MHz channels alleviating much of the congestion problem at the expense of cost and using a higher frequency.
http://en.wikipedia.org/wiki/List_of_WLAN_channels
for each additional hop away you halve the bandwidth and double the latency.
I don't follow you - why couldn't remote nodes continue to pass along a chunk of information at a continuous speed? (i.e. bucket brigade).
You're supposing a half-duplex medium, but even in that case, the medium does not have to remain locked up in the local vicinity just because packets which originated there are still being propagated further down the chain. Indeed for a mesh of any appreciable size you would need to implement local CSMACD and collision recovery in some shape or form anyway, to deal with the case where two transmitters on the mesh, out of range of each other, send overlapping transmissions which eventually collide at some intermediary node.
And doubling the latency is only true _for and half duplex medium_. If you had transceivers which could transmit on one channel while receiving on another, then you could pass through the data with negligible added delay - assuming you don't store the whole packet before forwarding.
All the assumptions you've made here seem to be of a rather specific and primitive mesh technology. They are certainly not true of meshes in general. From a theoretical standpoint it is not hard to see how those limitations could be overcome with slightly more sophisticated technology.
One of our competitors trademarked the term "hypothesis". From now on, we will call them "boneheaded ideas".
In fact the rate at which latency increases and bandwidth decreases in a mesh network has been shown to be 1.68^N where N is the number of hops (Piyush Gupta, Robert Gray, P.R. Kumar An Experimental Law for Ad Hoc Networks, May 2001). I have measured this myself in an 8 hop half duplex wireless network.
That is only for data being sent straight through the mesh. When clients connect directly to the mesh nodes on the same radio that is used for sending and receiving mesh traffic then there is even less radio time to go around and things slow down even more.
There is a company called Mesh Dynamics which sells mesh nodes with three radios: one for send, one for receive and another for clients to connect to.
I recall they have modified or extended in some way the linux kernel to have one radio dedicated for sending and another for receiving. This is something I am interested in doing as well but haven't seen any kernel modules or extensions to allow it thus far.
I just now put an entry in my hosts file:
127.0.0.1 ad.doubleclick.net
Hopefully that will do the trick.
The halfing effect will certainly occur if you only use single radio systems. If you use multiple radio systems that run on different frequency ranges, you can do much better. A good meshing algorithm will avoid the halfing effect by properly using the alternate radios available. If the OLPC had multiple radios, it would do much better as a meshing node.
There are some differences between a many hop city mesh and what I would call a dense mesh. City meshes usually are trying to cover as much space with as little equipment in order to obtain good enough performance. In my idea of a dense mesh each node can see many if not nearly all of the other nodes. The biggest problem becomes route selection when you have lots of very close and similar routes. This can be solved with static configurations but that fails when the mesh nodes themselves are mobile. For that you need a good dynamic route selection algorithm. This kind of dense meshes can occur with mesh nodes in vehicles that are moving around each other.
I wish the OLPC was truely open so that we could understand the route selection and 802.11 rate selection algorithms used. These two algorithms can be a significant effect on mesh performance.
-- soldack
Inaccessibly placed solar panels (such as the one suggested placed atop trees) are basically worthless in temperate environments due to snow. You can even have an excessive number of such mesh nodes distributed in tree-tops and the redundancy will essentially do you no good due to the fact that snow blankets them all -- and right when you most need communications for the rural areas which are, then, snowed-in.
Seastead this.
I've been working with some people trying to come up with a cheap solar mesh router. now this guy comes along and says it can be done for $35, so no one is going to pay more. he obviously has no idea of what is involved in making a solar mesh router that is the least bit reliable. but people aren't going to know that. They're just going to have heard this magic price of $35, and not be willing to pay more.
olpc was supposed to be an open-source laptop for $100. yea, whatever happened to that? closed wifi hardware, microsoft involved, screwed around by intel, the price still high, the computer not available.
you're just salting the earth so nothing else can grow. We had been hoping to create a mesh router that could be sold for $100, but we can't even get the components for that cost. I don't believe you'll ever see a $35 solar mesh router. Now, since you've damaged our already struggling project, you'll likely never see a $100 one, either.
So, like... do I type that in my Playstation joystick? I kinda can't seem to locate the Pi button, it must be the circle one.
Oh, and while we are at it the Playstation 3 is SO much better then the Wii!
alias possession='chmod 666 satan && ls
I worked on an early product line that did mesh radio networking as a broadband alternative to cable and DSL. It was a good solution for where cable and DSL didn't reach, and was better than line-of-sight wireless. We needed one big name ISP to get behind it to keep us viable, but all we had were lots of really tiny ISPs so we were eventually shut down. There was some attempt to get the mesh technology into other realms before that was cancelled too.
Then a year or so later I started seeing reports about all the competitors coming out with meshed solutions, including some of the same research ideas that we had proposed (though usually with WiFi). It's interesting to see if the OLPC goes with this, and I'd like to see how it works out. We had used mesh to get around line of sight problems like hills and tall buildings that the Australian outback won't have to deal with, but the concept is extremely similar and adaptable to other locales that aren't pre-wired.
Virtually everyone could have one. So if each person managed their own would it really be that bad? How hard would it be to write some software that would connect to your specific repeater to test that its working? If you can run this off a solar panel at this point then its not using very much juice. Its cheaper to spend $35 one time and take some ownership in this new interweb thingy then pay $40 a month for high speed access that is throttled to death by our communications overlords.
WiFi is really a bad medium for meshed networks, and isn't very good for sharing bandwidth (like ethernet it's CSMA). WiFi mobility is also difficult. But it's the best choice for inexpensive solutions that need off-the-shelf parts and not a lot of development under the hood. Our meshed broadband solution had used a custom TDMA MAC, but our industrial world target user who wants to play Warcraft or use Skype is very different from the OLPC's target user.
Mesh is a fairly new technology, ...
Not really. The OLPC's mesh sounds suspiciously like the "chaosnet" that has been in use on the MIT campus since some time in the 1970s. That, too, was decentralized, with each node forwarding packets for its neighbors. I've wondered whether it was sheer coincidence that the OLPC's mesh was developed at MIT. But so far, I haven't run across any comments on the topic.
It isn't especially surprising that this sort of technology hasn't been available commercially. It does sorta shoot down the central control that the commercial guys like to have. It's a lot easier to keep your customers in line (and cooperate with the feds' demands) if all traffic has to go through a single chokepoint, and you have control of that chokepoint.
It'll be interesting to see if the US government and comm companies permit this sort of network to be deployed widely within the US.
Those who do study history are doomed to stand helplessly by while everyone else repeats it.
Trouble is, you need sufficient production volumes to get cheap mesh networking. Current consumer-level wireless chips (i.e., 802.11a/b/g/n) are less than well suited for mesh.
If J.K.R wrote Windows: Puteulanus fenestra mortalis!
The boxes for the mesh radios are BLACK! Totally stupid decision, as they do not reflect sunlight and allow the guts of the box to heat up to levels which will either degrade performance or cause the radios to malfunction completely.
How do I know this? Let's just say I've learned from personal experience.
Antarctica?
Semi-automatic amateur armchair Australian philosopher; conjecture ready at any moment...
I stand corrected!
While black is the best color to radiate heat, it's also the best color to absorb it. When trying to perform passive heat dissipation, there are no easy answers. Your tools are conduction, convection, radiation. Heat flows from hot to cold. The sun is hot and usually whenever you care, the outside is hot too. Have fun!