Anyone looked at purchasing bulk bandwidth and distributing it over wifi to homes for profit? Basically, being your own ISP?

In my neighborhood, there are ~65 homes and we have lights (meaning 120V avail) on 6 foot 4'x4's scattered around the neighborhood which would be perfect for installing a mesh of routers to blanket the neighborhood in signal.

At this scale the administration costs would probably quickly overwhelm any costs savings gained (if they are even possible) but I've spoken with a neighbor who is an older IT professional that simply can't get hired in this economy that would likely love to manage the system for next to nothing.

This is kind of an open thread. But reading this reminded me of some thoughts I had about this before

http://arstechnica.com/tech-policy/news/2010/10/profs-bring-free-white-space-broadband-to-working-class-houston.ars

10/11/2010 10:14:27 AM

It will only take one dickhead to ruin the bandwidth for everyone else. Plus, how would you secure each home?

10/11/2010 10:35:33 AM

This would be a neighborhood system and would be managed in a way similar to your current ISP manages your traffic with throttling when necessary. And what do you mean secure each home? Since the neighborhood is small enough it's a simple matter for the system admin (and anyone interested in the operation of the system) to go into each home that needs the help and set the system up with the network key.

10/11/2010 10:50:21 AM

i thought WiMax was suppose to take off?

10/11/2010 11:29:47 AM

4G is WiMax and it is taking off.

10/11/2010 11:32:09 AM

ah did not know that.

10/11/2010 11:33:29 AM

^^^^ They're still all on the same network. A "secure" network isn't very secure with 65-150 strangers on it imho.

10/11/2010 11:57:12 AM

^^^4G is Wimax & LTE technically.

[Edited on October 11, 2010 at 11:57 AM. Reason : carrots]

10/11/2010 11:57:16 AM

I have experience with being an ISP.

For mesh wi-fi, I haven't done it but I've read good things about the Meraki APs, they are supposed to work well in a mesh environment. Used to be cheaper but I see you can buy them for $200/ea

Broadbandreports.com has a forum dedicated to wireless/micro ISP stuff, you could find some resources there.

If you already have cable and DSL in the neighborhood I doubt you would get most of those 65 homes to switch to you. They're going to be buying phone/tv from another company and they'll keep getting hit with marketing and ads encouraging them to bundle and get their internet from them.

Problems can kill performance if you don't have someone with good networking knowledge set this up and maintain it.

6ft pole sounds kind of low to the ground. A truck could park in front of one of your APs.

The wi-fi bands are subject to a lot of interference which may be a problem for you.

10/11/2010 2:33:39 PM

High frequency (2.4GHz or 5GHz) wireless doesn't have the capability to penetrate much. You'd basically need an AP for every 1 or 2 houses. But if you use any other frequency then you're going to have to deal with regulatory issues. I thought about this a while ago for my neighborhood when I found out how much TWC charges for business class internet access. I found out from Level 3 that the bandwidth costs like 2-5% of what TWC charges, but obviously you need your own infrastructure, which would run well into 5 figures, possibly 6 for my neighborhood.

Getting broadband the "last mile" from the backbone to your home is a pretty big investment unless you're sitting on top of it.

10/11/2010 3:17:27 PM

Ironic that I too have been looking into this recently. Except we have access to a factory roof and a water tower for distribution

At first I was considering the motorola Canopy system (I think it's lost the moniker and just uses PMP for point-to-multipoint now). The access points are *pricey*, but they seem to pack the right features for our application (we would use a tower on the roof relaying to the water tower; and probably a long range repeater on a telephone pole in the two neighborhoods we'd like to service.)

However, you've got to really make your case to get marketshare. Good friend is the admin for the deployment of city-wide fiber optics in Stokesdale. They're providing 50Mbit down, 15 Mbit up for $55/month and are finally starting to get some marketshare. They're hovering around the 25% mark coming up on their first anniversary - but that's 5x the bandwidth for basically equal money with the competition.

Of course to do it right you need multiple backhaul points and redundancy - which both cost money. And you have to bank that people aren't douchebags and run bittorrent on your fast network 24/7 (or figure out how to throttle it - which I almost refuse to do. I want an open network.) Customer service, answering emails and phone, blah blah blah - definitely a touch market.

10/11/2010 11:21:22 PM

I realize this thread is about being your own ISP, which I have not, but hear me out...

[Edited on October 11, 2010 at 11:56 PM. Reason : ]

10/11/2010 11:55:22 PM

hahahaha

10/12/2010 12:22:27 AM

I worked for a company in highschool and part of college that was able to implement a Motorola Canopy network fairly well. They already had an established dialup customer base though. Started with 2 bonded T1s and later had 10Mb fiber from TWC. No idea what they have now.

10/12/2010 8:40:02 AM

Quote :

"motorola Canopy system"

A company here in wilson uses this to on several water towers/hospitals here in town. Bouncing signals, etc.

They put a sign up on the fence around the water tower (CALL US FOR INTERNET 252-XXX-XXXX). Most of these water towers are in the country side where time warner is scarce. I think they reselling spirit bandwidth. I also want to say they've struck a deal with some of the smaller towns surrounding wilson to allow them use of the water towers free of charge but they had to supply their municipalities with internet access.

I've logged into one of the canopy system satellites mounted on a building. They come with a lot of debug tools and report tools to let you know you are getting a signal.




[Edited on October 12, 2010 at 9:03 AM. Reason : s]

10/12/2010 9:01:19 AM

I'm actually doing something similar to this in another country. You will need to use a licensed band for this to work in the US. There will simply be too much interference on 2.4ghz to have anything that could compete with the reliability and speed of a fixed line. To get into the licensed spectrum space, you'll need a much larger customer base than ~65 homes to offset the cost.

10/12/2010 6:59:48 PM

Thats why I included the ars link. I assume we'll eventually get a new 802.11 standard using this new space. That should solve the penetration issues. To address another poster though, in my 'hood these 6 ft poles are actually fairly numerous and spaced at about every other home. I didn't count them, but I'd guess there are 15-20 for the 65 homes.

10/12/2010 10:03:02 PM