Select Star Logo
December 4, 2017

Community Based Internet Projects: IoT Solutions

Generic Placeholder for Profile Picture
December 4, 2017
Stephen Goldberg
CEO & Co-Founder of HarperDB

Table of Contents

After my last blog post, Why I am Excited about the Repeal of Net Neutrality, I received a lot of great feedback from some very smart people.  My post was essentially about how I see the current infrastructure of the internet being replaced by mesh networks which are decentralized, and would allow for true internet freedom.   Most of the feedback I got was, that while this is an interesting idea, it is too complicated and too expensive.  I disagree.  

Throughout my career I have been known for pushing forward many Don Quixote-esque projects so I thought I would take a look at how I would tackle this project.  I decided that for this blog post, I would examine what it would take to build and maintain a mesh network near and around my home.  I live in Larimer County Colorado with a population of 333,577 spread across 2,634 square miles.  The goal of this project would be to provide the option for community based internet to 97,164 households in Larimer County. I see three main issues to be addressed; implementation and maintenance team, hardware, and tier one connectivity.  

Disclaimer: I am not a network engineer so forgive me if any of my assumptions are incorrect. I am mainly interested in exploring this concept in the hopes that I will get feedback from the community as well as potentially see others adopt my ideas.

Local IoT Projects

I have been thinking about this a tremendous amount, and I think the only model that will work is to have local nonprofits be the driving force behind mesh networks.  The incentives to sell data, and exploit internet usage are too high to trust private companies or government agencies. This does not mean that I do not see a role to be played by government and private industry in the new internet; however, I think the ultimate control of these networks need to be handled by local nonprofit entities.    

Look at the Open Source community; nonprofit organizations like the Linux Foundation have been the driving force behind keeping Open Source truly open and free.  Companies like Red Hat, Intel, Google, and many others in turn have supported those organizations because they see the value in the innovation that is driven by that community as well as the excellent brand recognition they gain from supporting Open Source.  

My hope is that we would see something similar with community based internet, where a Cooperative would handle the management of the mesh network while government and industry would support it through partnerships, grants, and donations.  

The Foundation

In order for Larimer County Colorado to have a mesh network that provided Internet to all its homes they would need to form a non-profit Cooperative, lets call it the Larimer County Community Technology Project for the purposes of this blog.  This concept is already in place here as well as other places across the United States for utilities.  I get my power from a Cooperative called Poudre Valley REA.  This doesn’t mean it’s free, it means that home owners who get their power from the Coop are also members of the Coop.  Internet has become a utility like all the others.  Lets be honest, most of us need it to survive at this point.  

I see the cost breaking down into these categories:

  • Initial hardware costs (edge devices, access points, etc.)
  • Construction costs
  • Volunteer labor for initial implementation
  • Paid labor for on-going maintenance
  • On-going cost for tier one access point

My thought process is that a volunteer project team would raise funds via donations as well as potentially local, state, and federal government grants programs in order to fund the initial implementation.  

After implementation, homes that wanted to access internet via the project would pay a small monthly fee to the coop for access.  I project that this fee would be significantly smaller than what we pay today, as I will discuss at the end.  The cost of maintaining a small staff of full time network engineers, replacing hardware, and paying for Tier 1 connections would be distributed across coop members and hopefully subsidized by government grants and corporate donations.  

Larimer County has four major population centers: Fort Collins, Loveland, Estes Park, and Berthoud.  For the most part the county is a fairly rural area with homes being spread out on multi-acre properties.  Surrounding major population centers are more densely packed.  

I will get into the hardware and backbone in a bit, but basically this would be a multi-phased project.  We would place Tier 1 access points in Fort Collins, Loveland, and Berthoud, in that order.  As funding became available via grants and with members joining the coop, we would begin spreading the radius of the project out further and further.  Obviously we will have some signal degradation as a result.   That said, many folks who live west of Fort Collins and Loveland currently, either have no internet or pay an exorbitant amount for terrible satellite internet, so I think that this model would be a significant improvement in quality and cost.  

IoT Gateways and Access Points

As mentioned above I think the project would need 3 to four Tier 1 backbone connections.  How would we possibly get those in such a rural place you might ask?  The construction costs would be astronomical you will decry!  Are you sure?  Or is that what some ISPs might want you to believe?  

Below is a map from Zayo’s website that indicates where their existing dark fiber runs in Larimer County.  

The blue lines running through Loveland and Fort Collins are dark fiber.  As a result you can see from the above that getting a direct access point in Fort Collins, Loveland, and Berthoud should be pretty straightforward as there is fiber running directly through those towns/cities already.   We would then use these access points as foundation gateways and for the mesh network.  I have some experience in buying Tier 1 connections and they are expensive.  I have experienced costs of $1350/month for a dedicated 1GB static connection.  Assume that the connectivity we needed would be $10,000 a month per location and is it likely the providers charge us twice that because nothing is ever what you expect.  I’ll set the total cost to $20,000 per access point per month.  I might be grossly overestimating this, but will keep it conservative, better to air on the side of caution.  We would need 3 access points, so $60,000/month total or $720,000 a year. For estimating purposes lets say they charge us $50,000 per access point for the initial connection.  I have witnessed quotes of $60,000 in San Francisco to run fiber into an office building, so I think this estimate is close.

These numbers are all entirely made up, but I think they are probably in the ballpark from my experience.  The next requirement would be a local business or homeowner to allow the connection of the tier one fiber to their location at each access point, then install hardware on their roof.  These businesses or homeowners are to receive an incentive, for example free internet as well as the benefit of the best connection in town!  

Infrastructure:  IoT Devices

With the access points installed, we can then move on to getting internet to my home which is 5.2 miles from downtown Loveland.  To do this, a daisy chain is constructed of heavy duty outdoor omni-channel and directional WiFi range extenders.  The extenders cost about $600 retail and we would need an antenna, which cost about $80.  These kits would cover a one-mile radius in the case of the omni-channel antenna’s, and a one-mile distance in the case of the directional.  Estimated total cost of these kits would be $700.  Again, to err on the side of caution, lets more than double the estimate to $1500 to cover every square mile of Larimer County.   That’s 2634 square miles so about $4 million.   Assume that the entire network needs to be replaced every 4 years, costing an additional $1 million a year in maintenance.    

Costs

Having all the needed components, take a look at them all together:

One time costs

Total: $4,150,000

Recurring costs

Total: $2,798,000

There is an approximate upfront cost of about $4 million and an annual cost of almost $3 million. It sounds like a lot doesn’t it?  It does if you are going to front the bill yourself or if you are planning on rolling it all out at once.    

On the other hand, there are 60,503 houses in Fort Collins and 28,557 houses in Loveland.   Imagine phase 1.0 of the project is a single access point in Loveland and one in Fort Collins and set phase 1.0 estimates at $100k in upfront costs for fiber.  Covering a 20-mile radius around each city – that is $60k in WiFi hardware.  Assume 10% of the population of both cities register, which supplies 8,908 households.  Upfront costs would be approximately $160k.  Distributing that 160k across those households is $18/per house.  Our monthly cost would we at $5 per house.  Charging $30 per house would generate roughly $62,356 a month and portions of that could be used to expand the network every month, so it is doable.  Getting initial donations and government grants prior to building for probably $250k, would be ideal and a sufficient amount to get started.  This would help cover the build out and several months of connectivity, while building up to 10% of the population.    

Now imagine we ultimately got the entire network built out as I have laid out above, and we got 25% of the households for all of Larimer County which amounts to 24,250 households.   Our initial costs are $4.1 million and our annual costs are $2.8 million.  If we amortized the $4.1 million over 3 years, and rolled that into our $2.8 million, it would amount to a total of $4.2 million per year.  That comes to about $14.00/month/house, which is half of what we were charging before.  We could actually lower the rates over time as more and more people came online, as our primary goal would not be profit, but rather freedom.    

Challenges

I am not trying to say that this would be without challenges.  As I mentioned above, I am not a Network Engineer.  Getting cooperation from the local governments would be difficult as well as finding people to agree to put routers on their homes.  Convincing folks to agree in areas where property rights are a pretty big deal would be really hard.  Imagine an area where someone owns 1,000 acres between the west side of the county and the access point, and there is no way to go west without going through their property.  What if they will not allow range extenders on their property?

Range would be an issue also. The furthest point of Larimer County from where we would place an access point is from the Northwest Corner of the state to Fort Collins, which is about 64 miles.  Daisy chaining wifi across 64 miles would certainly result in a pretty dramatic signal degradation. There are probably ways to improve this with the use of directional antennas instead of omni-channel and other IoT Solutions, but still it would be problematic.  

I imagine that 25,000+ users on three access points would also strain the system dramatically.Getting cooperation from a fiber provider when so much of their business is dependent on the existing Internet business model would also be problematic.    

Finally, security would be a major concern as well and something that would need to be watched closely.  

Conclusion

Despite the numerous challenges in getting a project off the ground like the one outlined above, I believe it is possible.  Not only in Larimer County, but also all over the United States.  I believe people will start to tackle these problems one by one and with grit and determination and launch some really incredible projects.  I think we are at a point in the evolution of technology where solutions can be found for all of the technical problems I mentioned above, or will come about shortly. If implementations are handled in a measured and phased approach they will have greater chances for success.  Phases will be the key to getting the network hardened and generating enough cash flow to continue to pay for its expansion.  For the technical problems that cannot be immediately solved or for the non-technical problems, compromise will be necessary.  In a majority of cases, compromise is probably significantly better than the alternative.  

These are examples and what the internet looks like in Portugal without Net Neutrality.

Cost of net neutrality

Aside from freedom of outside influence, there are other incredible benefits of a mesh networked community based internet.  Imagine being able to connect to WiFi in every park, in every restaurant, in every area of your entire town with the same personalized username and password?  Your connection could be as private, secure, and fast as if you were in your living room.  No more goofy WAP pages at the Starbucks and potentially even no more cell phone bill.   I think companies will also be incentivized to support the proposed model.  Imagine if autonomous vehicles could connect to WiFi everywhere they went.  This would hugely benefit them, and as a result might encourage them to support an effort such as this.  I could go on.  In the end, I hope to see these projects begin to take shape over the next few years, and I strongly believe they will.  

If you have feedback, comments, or questions please feel free to leave them below or contact us at hello@harperdb.io.

While you're here, learn about HarperDB, a breakthrough development platform with a database, applications, and streaming engine in one unified solution.

Check out HarperDB