Our Internet

This post is inspired by an article I just read in Linux Journal. It is Doc Searls interviewing Bob Frankston and is titled Beyond Telecom. I am not posting a link as the article isn't available on-line for free yet.

A good starting point, however, is on Bob's site. This is one of many and I also highly recommend his Beyond Network Nutrality article.

This quote from the first article will help set the direction.

Today’s Internet is treated just like another service in the image of the phone networks or even another channel on your Broadband Set Top Box. And to a large extent that image is reinforced by the central authority of ICANN as the dispenser of identifiers and identities.

But this isn’t in the spirit of the Internet which is defined by the user/developers outside the network. The Internet is just a name for the community of those cooperating and using common protocols. Instead of viewing the Internet as something we connect to we should view it as radiating from us and our devices. Our home networks reach out through community networks and beyond finding paths that work.

Put another way, he is saying the purpose of the infrastructure is to transport bits and nothing more. But, the business models of telcos and cable companies have them charging for "a service" instead. His example in his interview with Doc is that a bit in an SMS message costs much more to the consumer than a bit in a video stream but there is no difference in the two as far as the network is concerned.

I bring this up here because Nicaragua has a lot less existing infrastructure than the U.S. in terms of moving bits. It is growing rapidly but maybe there is still time to think about what Bob is suggesting. It boils down to free local connectivity, cacheing at a local level, community networks, ... This may all sound very complicated and technical (and, on one level it is) but the socio-political aspect of it is very simple.

Let's look at what this could mean for a community. I will pick the ficticious city of Netville, Nicaragua to talk about it. Netville has a population of 30,000 people and (for convenience) is located on the Pan American highway. It currently has only dial-up Internet access.

Step one is for Netville to get a whole bunch of WiFi access points and "scatter them around". That is, build a local network where all of the residents have access to a WiFi access point and all the access points talk to each other. This immediately means anyone with a computer and a WiFi card can talk to anyone else with the same thing in Netville. If Netville high shcool decides to make homework assignments available on-line, all the students can now have access. Same for government, ...

Now, while this may not seem unique, what you have is local communications with virtually no network costs. Contrast this with Estelí today. I have Enitel DSL. If I want to send a message to someone across the street that has a cable modem, the message has to go to the Enitel office (and I am not sure if that means here or in Managua) get routed to the gateway from Enitel to Estesa and then delivered over Estesa's network.

Estesa and Enitel want you to pay for sending that message because it used their facilities. But, that was only because there wasn't a local way to deliver the data. That is, their design forced an inefficiency.

I don't want to make this too complicated but let's just toss in video (let's say the 6PM news) over your "local network". If you want to watch the same show as 500 others in Netville, you just need to deliver one copy of the news to the Netville local network server. It is cached and routed to all who want to see it.

That should be enough to get some people interested and others lost. In social-political terms, it means looking for a local solution and then growing that solution to connect to other resources. It seems like something that could be done, would be good for Nicaragua/Nicaraguan communities and would even appeal to some organizations looking to make grants. Much like the grants for the sewer and water projects in Estelí, this is infrastructure. Up-front costs are much larger than operating costs so it doesn't build that dependency cycle.

Am I crazy?

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.

ad-hoc

If your PC and the guy sitting across the street in Esteli have a WiFi card set to ad-hoc mode then you could send him a message directly and by pass the Central Office. A few of these around the neighborhood and you have a mesh network and they could all use your DSL for the internet connection.

Guess I'm forgetting a way to route the message, but that would be easy if you're using Linux.

routing

Routing actually is the issue. But, you are absolutely right—that is what I am suggesting. If the community (that could be an organization specifically doing the project or it could be the city government) decided to get involved they could add some access points with internet connectivity and you are good to go.

The best way to do this is to avoid the things done wrong in operating systems. While the OLPC is designed to mesh network (and, you are right, it is easy to implement with Linux) the majority of the systems would require tweaking. You would also be dependent on someone having their system on. I see the solution as two types of access points.

The first type is as mentioned above—organization or city-supplied with connection to the internet. One way to do this (and there are units commercially available that just do this out of the box) is one with two radios. The 802.11a radio is for the backbone connection to the Internet. The other radio is a normal WiFi 2.4GHz radio that users connect to.

The other access points would just be normal WiFi units. Much like having a DSL or cable modem, this unit would go in your home and you would connect to it via an Ethernet cable. It would give you network access and also act as a repeater for others.

The routing is the trick. There is an in process standard on this, 802.11s. If you are not a geek, this Wikipedia page is probably a better place to start.

I am Definitely

Not a geek. I wish I knew mostly what you are talking about although I realise it's probably music to someones ears, Just what are you proposing?

mesh is a cool topology..

we never had an opportunity do do any hands on with it when I was going to geek school. Things were geared towards Microsoft's NT vision. Wireless was that "unsecured option that had little real world application", bless their cold and steely little hearts.

With 802.11s isn't there power that goes to the WiFi card even if the CPU is powered down ? Wouldn't this open node act as a repeater of sorts? And how is key authentication handled without a trusted server?

The IEEE working group has and online class for rusty old geeks that want to catch up on this topology but it seems to be down at the moment;

http://grouper.ieee.org/groups/802/11/

onward through the fog...

-Doug

If you're not part of the solution, you're part of the precipitate

New to me too

I am no expert. Willy and I talked about "some sort of local connectivity" when we both lived in Costa Rica but we never found "the solution". Since then, mesh has become the answer.

Keeping nodes powered up was why I suggested an external AP with an Ethernet connection to the computer. We are talking very little power—on the order of maybe 1W average—to keep the AP on.

As for authentication, it seems unnecessary. The transport is what you offer for free. Specific services (once you reach where they live) still costs and has security associated with it. This "transport is free" piece of the picture is hard to swallow at first but it ultimately does make sense.

When you start up, you have lots of local bandwidth (it actually grows automatically as the network grows). The critical resource is your connectivity from the local mesh to the Internet. But, the first thing this does is encourage local cacheing servers and as much local service as possible. If the community internet connection is saturated, you have the following options:

  1. If you/your business depends on a certain amount of internet connectivity, you can buy that outside this system. If you are a good community citizen you can then share what you don't use with the community. This just becomes a QoS routing issue.
  2. You can cache more on the local network. For example, cacheing news programs locally if they are popularly listened to/watched.
  3. Get the community to add more (paid for) connectivity to the internet.

What's important here is that it can always be a local decision to do what is best for the community rather than being forced by "the big guy" to do what is best for their business model.

I am pretty interested in seriously developing this idea. That is, to get together with a few people, toss around the technical issues and write up a proposal to do a test implementation somewhere in Nicaragua. That really means getting the government (certainly TELCOR and also some local government) on board.

Done right, it seems that the important pieces could happen:

  1. A grant from someone for the hardware and at least the salaries of the technician(s) needed
  2. Generic permissions from TELCOR for the network free of license fees
  3. A bit of pressure (probably from government) to get those that provide Internet connectivity to accomidate the project with, for example, a non-commercial priced connection (DSL, ATM or cable).

    If you have any interest in being involved, PM me or email me at nicafyl at gmail dot com. If I get a reasonable number of people interested, I will set up a mailing list so we can "talk".

Cost

I'm no expert either, but I think cost would be a huge factor for the network. On e-bay for less than $30 you can buy a Linksys wireless router.

The one I have is about three or four years old and I up loaded or flashed the EPROM memory with some free firmware from Sveasoft replacing the manufacturer's official software. This enables me to boost the transmission power of the router from about 30mw to about 250mw and get more range. I can also telnet into it for easy management. It is basically a linux computer that can transmit and receive messages for under $30.

With a little more work and a Pringles potatoes chip can, you can make a directional Yagi antenna and boost the range even more(point it at the Geek Ranch). One of these on every city block and you have connectivity all around the city. In a small town like Netville, you wouldn’t need many.

You could even sell subscriptions and limit access by IP or MAC address to spread the cost around.

open source..

firmware from ;

http://openwrt.org/

sounds cheaper than Sveasoft:

"Sveasoft firmware and support is available via an annual $25 USD subscription"

The directional antenna sounds interesting, and cheap, any estimates on the range?

I'm new to open source, just got fed up with being an unpaid beta tester for Microsoft, and want to do without them as much as possible.

-Doug

If you're not part of the solution, you're part of the precipitate

openwrt

The openwrt.org firmware looks good and is probably more current. It was some years ago I flashed my router and never got support. As long as I have power, I'm good to go. It has a GUI to tweak the signal and security which openwrt probably has too.

The yagi antenna is a popular homebrew project. Seems to me I saw some site where a guy got a 10 mile line of site connection. One guy started going to the suppermarket with a tape measure looking for the perfect size can for the wave length.

OLPC Mesh Networking

Let me toss another article, this one about One Laptop Per Child and Mesh Networking into the mix here. The article explains in pretty basic terms what a mesh network is, why it matters and even addresses energy issues.

electricity comes first

FYL, the problem is the traditional tendency towards monopoly of a given service or commodity. The Pellas family has done a fine job slowly accumulating their wannabe chaebol or keiretsu (conglomerates). This is the same behavior that the Somozas engaged in.

One of the things that drives me crazy about Nicaragua since 1990 is the pricing of goods and services without taking into account the average man's purchasing power. We all know that internet in Nicaragua remains a luxury. The average person cannot afford $60-70 per month. And what do you get for that price??, speeds that are 1/4 of those in the U.S.

If prices were fixed re purchasing power parity, that internet would cost less than $20 per month

Add to that the dismal state of the nicaraguan electrical grid and you will see where your netville dreams falls apart. Computing by necessity assumes a reliable and stable electrical grid.

No sizeable investment will be made in the electrical sector. You need first to undo the piñatero culture of theft of services, and that has been deeply rooted since 1979.

$60-70/mo, is that the cost of internet in Esteli?

And what does the phone to connect cost? Is this DSL, or what? If not, is DSL available around Esteli?

Thanks!

Wu Wei

Esteli Connectivity

DSL is from $20 to $55 for homes; lots more for commercial connection. The $55 is 1Mb down/128Kb up. Dial-up is available but DSL or cable quickly is a better choice as there is no such thing as unmetered phone calls in Nicaragua.

Proposed Modification

You brought up two valid concerns. The first, the price of connectivity is one of the main things addressed in the "community connectivity" plan.

Now, on to electricity. First, it at least where I live, it has been very reliable for the last year. But, the grid infrastructure sucks resulting in huge losses and, as a result, higher costs. So, it does need to be addressed.

I see no reason why distributed energy sources to run the distributed network can't be part of the plan. Netville, for example, can have their own solar/wind/whatever power source. It can be plugged into the initial grant requirements.

As for individual users, that can also be addressed. The One Laptop Per Child project, besides offering inexpensive computers, even offers a hand-crank generator option.

The main thing that Bob Frankston is talking about is moving control from "them" to "us" meaning to those who actually use the connectivity. Then, you can actually start solving the real problems without bumping up against business models of those that currently control electricity/communications/...