Topic: Decentralized phone service (Read 10344 times)

test

Wouldn't be particularly useful in a real emergency anyway.  A single repeater station doesn't really have enough useful range for that.  However, I've been following a new wireless tech called Dash7 that I'd love to have on my phone, that has a functional range of 2km at 6 milliwatts, as well as an automatic 'ad-hoc' mesh network topology.  Couldn't support speech or any high bandwidth data of any sort, but it looks like the Jabber protocol is a beautiful fit, and would work great in an emergency with some GPS code extensions to the 'presence' part of the protocol.  It also looks like it would be a good way to distribute Bitcoin TX (transaction notices) to nearby mobile devices in a city that Bitcoin were used to buy things IRL, should that ever happen.  If Dash7 becomes common on smartphones, someone will write a phone to phone texting program eventually, and Dash7 would become the texting version of Citizens' Band radio.  It may be noisy, often slow, and prone to interference; but it would be free and not subject to a need for infrastructure support.

At burning man they setup a GSM station with a VoIP bridge of some sort.

Were mobile users able to call each other, completely isolated from the countries network in that example?

I think really cell base stations should have an overload feature where any overload during a crisis means cell users can at least communicate locally, perhaps if only by SMS.

Don't really know enough about GSM on how that could be setup though... automatic if demand reaches a limit would be good

Wow.  I don't know how this thread slipped by me.  Hugolp, you're explanations were very helpful. Thank you!

I would be very interested in trying to build a physical nationwide network with something like this.  If done simply enough, we could probably get mainstream (eg, savvy, but non-hacker) people on board.

Is anyone looking into this further?

Hmm. I am thinking of getting all the hackerspace all over the country to act as supernodes. Hackerspace are learning and community centers for...hackers.

This network has been created differently than the potato mesh idea that you posted. There are nodes and super nodes. They have different functions. Basically the node connects (usually) to a super node, and the super nodes connect between themselves. The reason for this is mainly hardware and easier routing. The nodes are basically simple devices with one antenna that connects to a super node, like the one I just linked previously. Super nodes have different antennas, some directional to connect to other supernodes, and other omni-directional (or partially) so the nodes can connect. The reason for this is that directional antenas only radiate in one direction but get farther away and with better bandwith, so they are used to connect super-nodes. Then they have one or two (or more in some big super nodes) omni-directional that can not radiate km away but allows for local nodes to connect to the super node, because it radiates in all (or some) directions and covers an area. So you basically have the troncal connections between the super nodes and then the nodes connected to the super nodes. As the network started growing there was an effort to create redundant connections between the super nodes, for more bandwith and in case one super nodes goes down temporarely.

This is the reason why the supernodes (more expensive) are sometimes funded through a donation system, where the interested chip in.

The difference with what I have read of the mesh potato network is that with this system you can connect nodes farther away from each other, and the routing is more efficient, allowing for better lag and most probably better bandwith. I have my doubts on how a mesh network can behave in networks with a big number of nodes and covering a big territory, but if you want to cover a big area you need a lot of people adopting it. But this way, you can have a few people in different towns conected between each other, without the need of all the towns in the middle collaborating at the moment. The con is that you have to coordinate the IP addresses of the nodes. In the catalan network this is done through the webpage of the association. You create a node there (they use google maps to place the node) and then you get an IP address, that is usually dettermined by location and the supernode you are going to use, so the super nodes have an easier task routing packages.

If this kind of networks gets popular you can probably create a mixed system, where you keep the super nodes structure but connect to the super node using a local mesh network instead of a registered node. They are trying to do this here in one of the towns where guifi.net is more extended. Their idea is to use the local town government lighting and put the mesh nodes on top (basically because its elevated and already has electricity), so they create a mesh network that gives connectivity to the whole town, and even to your mobile phone or device anywhere in the town. But to do this there has to be a wide acceptance of the project, so I think starting with the node-supernode infrastructure works better initially.

And yes, those system where cheaper before (or my memory fails). I will try to look for the shop where I got the prices. Still 60-70$ one time purchase is afordable. In a few months you have already saved that much by not paying your intenet provider.

PS: Here there is even a optic fiber connection. The guys connected this way obviously dont live in a city where this would be difficult.

Wow, thanks Hugolp.

To clarify, a super node does the same thing as a regular, but with more antenna power and more max connections?

Can a network of only regular nodes work okay?

Did they really used to sell that thing for $40?

Nodes are pretty cheap. If you have someone with a super node near you, you can buy a ready to use device for 30-40 euros (http://shop.openrb.com/index.php?main_page=product_info&products_id=140, prices have changed). This devices come with the antenna and the circutis inside a sealed box with a ethernet connection (the electricity comes through the ethernet as well), so you just have to place the device in your roof, pointing to where the other node is and connect the ethernet cable. You have to make some configurations in the machine, that can easily explained and you are set to go. The box is sealed and can go through rain, storms, snow, etc... (although if you leave in a region with extream weather you might want to look at the specifications for max and minimum temperature). Some of this boxes have achieved connections of a few kilometers if they have direct vision between the nodes.

Building a super-node is more expensive, but not as much as you would think. The minimum bill for a decent super node is usually 300-400 euros. This includes a machine capable of handling 4 antennas, and one or two directional antennas to connect to other super nodes, and one or two omni-directional antennas (or similar) for the local nodes. Usually each town or region of a big city needs its own super node.

In Catalunya the network has grown without any type of central or initial funding. People just got together and started creating their nodes. At some point the webpage started a system of voluntary donations to create the super nodes, because at the end they were going to be used for the whole town. But if you get 10 to 20 people in a town to cheap in 20-30 euros you are good to go. This has been used also to create solar powered super nodes on top of mountains (imposible to get electricity there) to connect super nodes that did not have direct vision. You dont really need a lot of planning. Once the network starts growing then some coordination is needed, but is mainly configuration issues. In Catalunya they end up creating a non-profit association to handle this. This is just how they have done it in Catalunya. Maybe in other parts of the world other aproaches work better, but the system is not that expensive.

The software is already developed (linux based) and easy to configure. People in the community is very helping and if you need any type of help I am voluntaring to connect the two groups as translator for any help any USA people would need to start their own network.

What it usually works to attract people initially and start growing the network is for someone to offer a proxy to the internet. The proxy can have the p2p connections blocked (with the exception of bitcoin obviously) so the users dont abuse. A user with a good internet connection or even some bussiness can do this, without any cost (just the burden of having a part of the bandwith used by other people). Offering free (as in free beer) internet for only an initial payment of 30 dollars is a very good deal and a lot of people will be interseted. Once the network is growing, people can start using it to talk by phone and other stuff. In the catalan network there is people thinking about offering a commercial connection to the internet, that could be way cheaper since they dont need to deploy the infrastructure.

This is getting long. If any one interested has any more questions just shoot them, and I will even connect them with some of the more tech savy people around here if needed.

If the team that is engineering the devices does it well enough then networks can spontaneously start anywhere where enough friends live close, then it can grow through acquaintances or even just a for profit who resells them saying "there's a good network of them between 5th st. and Riverside dr. or whatever.

So it look like people agree that it is a good idea. However, it will need funding and a team to engineer such a project.

I was thinking about how you could start a network in a local area. If a person was willing to front a few dozen they could raffle them off at a county fair or something and distribute purchase info for losers. You'd be guaranteed the bones of a network if you got the right number/area.

I wonder how fax machines caught on. If I was the inventor or producer I think I'd ship out a dozen units to each fortune 500 company to jump start it.

This would be a big step towards anonimity and avoiding any government interference on bitcoin. Right now the government controls the internet through regulating the commercial providers, people's access point. But this is a private network, people just connecting to each other with wifi devices in their roofs. It would be impossible for government to control, without going home by home shutting down the devices. It is also a cheap way to have network connectivity and scape the government sponsored corporations that provide internet service right now. In the catalan network there is people providing proxies to access the internet.

Thanks...

There is an "English" link in the top left corner

I only understand it a little better once I translate it out of Catalan!

Is it similar to the potato? But they have 10k users?! At first I thought it was a computer program, but then I saw a device.

Interesting.

Maybe we could form a nationwide network within the US with bitcoin users?

You might be interested in this: http://guifi.net

http://www.linuxjournal.com/magazine/mesh-potato?page=0,0


This sounds very cool. A small network would probably just freely forward all calls/data. But I bet bitcoin would be perfect for prioritizing and incentivizing call relaying.