Hacker News new | past | comments | ask | show | jobs | submit login
Start Your Own ISP (startyourownisp.com)
635 points by mahathu on Aug 17, 2019 | hide | past | favorite | 95 comments

My company Althea (https://althea.net) is making open source router firmware that makes it easy to people to set up incentivized mesh networks in their communities. It allows routers to pay each other for bandwidth which means that everyone hosting a node earns money for the packets they forward.

This ends the incumbency problem by making bandwidth a commodity. Want a byte, buy a byte, switching to the best provider automatically. No software knowledge required, fully plug and play. (equipment still required, can't break the laws of RF. it's just regular WISP gear though)

Here's a talk I gave on exactly how the system is designed and implemented.



Huh, how do you deal with the problem of law enforcement assuming that an IP is a single subscriber? Even if law enforcement believed it up front, it still is a hassle to the users now dealing with law enforcement requests.

The traffic goes out through an exit node so the only thing that a gateway’s ISP sees is an encrypted tunnel.

Ah ok so it’s not anything different than the existing mesh tech world, just another implementation. Thanks, I misinterpreted the grandparent comment, it sounded more like leasing existing connections and not routing them to exit points.

We are leasing the connection used to route to the exit point. If you want to think about it like that.

In general our only 'mesh tech' innovation is the mandatory WireGuard encryption for all user traffic. Which as far as I know is a first. The rest is the billing and some ease of configuration.

Yeah, I didn’t mean to imply that you didn’t innovate. I just meant that the fundamental baseline for a mesh is the same.

Looks interesting. I looked over your presentation but I am still unclear about who owns/leases the exit nodes. Could you clearly define 'exit node' for me?


Think of it as an integrated VPN, anyone can own and operate one. They charge like any other node in the network so they can make money or at least cover costs.

The main issue is one of trust, we have left it up to the user what exit nodes to add to their system beyond a few we ship in the firmware run by ourselves.

Personally I think this is a point of weakness long term, we need to decide on criteria for including other exit servers. Trust is hard

How do we know what's in WireGuard?

I'm not sure how to parse this question. So I'll answer both interpretations.

WireGuard itself is open source, our firmware and software is open source. So if the question is about encryption/software integrity go take a look.

If it's about what traffic is routed through WireGuard the answer is all user traffic to the internet, no exceptions.

That doesn't eliminate liability. If you're knowingly trafficking illigal content you're also liable for it. Afaik the only exceptions are when the content is being moderated .

How would you knowingly transmit illegal activity if everything is encrypted?

> Afaik the only exceptions are when the content is being moderated

Depends on the type of illegal content. There is no established case law that protects you as a moderator of CP. Yes, LEO will not arrest you usually for it, but there is nothing that positively makes you safe as a moderator working to remove and report CP content.

Start your own ISP in US - there, I fixed the title for you.

I would love to start my own ISP in my country, except that dealing with my corrupt government and corrupt bureaucracy existing here would mean I would have to have triple the cash of official investment, because here without "greasing the wheels" nothing works, and all of that is done by discretely leaving an envelope with cash on the corner of the discussion table that nobody in your presence will touch it. Also the amount of cash itself is a mind game on its own. You must discreetly research the person to see how much money they take as bribe. Too much and news about you being a newbie will spread like wild fire and all your future discussions will drain you of even more cash in discreet envelopes. Too little money and your request will get denied making you face the same person handing over another discreet envelope, and you better have at least the correct sum in there. So waste time asking for this mind game, waste cash that you'll never be able to officially report as deductible and then start implementing the points in article. Welcome to Romania ¯\_(ツ)_/¯

Romania has 14.3M (1) Internet users for an active population of 9.2M (2)

As the same time, about 98% of the population has mobile phone coverage (which implies some form of data access) with terminals that can be had for 0 upfront cost and ~10€ per month subscription with very large traffic allowance. Telekom (previous Romtelecom) can offer DSL services in most villages where a phone exchange is present.

Corruption or not, its probably a bad idea to start a wireless Internet provider in Romania because it already has excellent commodity connectivity virtually everywhere.

1. https://www.internetworldstats.com/europa.htm 2. https://www.romania-insider.com/romania-active-population-do...

That is the situation for cities, yes. On villages is another dimension. Broadcast is almost non-existent and everybody is relying on mobile. The few that have Digi (that's the actual former Romtelecom) are a drop in a bucket. As corruption is rampart, those on villages that require constant high speed connection usually bribe some city hall workers to let them get a fiber from the city hall to their house. The big majority of people, like I said, rely on cell data by doing hot-spots with them, and that gives them most of the time 3G due to weakness of the signal. Romania has coverage over its territory, but just enough to allow you to call an ambulance or police, but to rely a country side business like a farm that sell its products over internet that's not happening. We all wait to see what happens with 5G, and see wtf is the outcome of the current trade war US/China and most importantly to see from where to get 5G equipment for infrastructure. Huwaei is currently the incontestable leader of 5G tech but with shots fired from US govt. against it, and due to Romania being in NATO and having a US military presence we can't really just say "you know what? we'll get 5G from Huwaei". On the other side we are part of EU and current movement is to kinda ignore Trump and the trade war (see EU stance vis-a-vis Iran issue) and most big telecom companies from France/Germany do have talk with Huwaei. So we, a small fish, are caught between a rock and a hard place...and we wait...and we started to fall behind.

That being said, a WISP for villages would definitely has its place in current situation.

You are rambling and the situation you are describing is nonsensical and not related to anything that is going on in Romania.

Telekom is the former Romtelecom.a quick Google search should clarify that for you.

Coverage in Romania is 98% mandated to all operators by law and regularly checked (http://www.ancom.org.ro/en/ancom-has-verified-the-coverage-o...)

Bribing a city Hall worker for fiber Internet connection is again made up. City halls and all other institutions in Romania are not connected to the regular Internet, the connectivity is provided by STS (Special Telecommunications Services) to a private government network. In remote villages this is provided over satellite or radio link. Regular Internet is purchased by the local authorities from local providers (as needed) and is the same service that would be available to anyone else.

If you would run a local farm, do not host your eCommerce site using the available bandwidth, do it in the cloud, that's what grown ups do.

Please provide sources for your claims and do start a WISP if you want to address the rest of the population 2-10%, spread out over a country the size of UK.

Sounds like the perfect place for a bribe transparency website to name and shame officals that take bribes.

The thing about name and shame websites is how easily they are gamed.

Also I wouldnt host any of it within the same country.

Ironic, because this kind of neighborhood ISP scheme was how Romania got a good foothold on internet connectivity in the 1990s, according to my Romanian friends who remember the period.

Yes, 90's were Wild West for Romania. We had no legislation hence a lot of stuff could and was done by ignoring the state. Those in charge were more busy selling stuff to fill their pockets then legislating. And all those neighborhood networks eventually merged so right now in Romania, on any given city, you can have a maximum of 3 ISP to chose from. We still do have good internet but we are starting to fall behind and soon I'm afraid I'll see here the same situation that is in US as in one ISP on any given zone and crappy internet speed.

I'm in Bucharest, and on UPS. Since I'm from the UK, frankly, the price and quality (speed, uptime, etc.) are a continuing marvel to me.

Weird, I'd imagine the ubiquitous and insanely cheap pole fiber would be a much bigger problem for WISPs in Romania than the corruption.

Seems like it would be very hard to make a profit unless you can find hundreds of customers close to each other who are willing to pay unusually high prices for internet connectivity.

Oh hello everyone! This is my website. Not sure why it hit HN today but cool too see it here! Thanks @mahathu!

I'm not at a computer tonight I'll run through later today or tomorrow and answer questions. My contact info is on the website as well.

Do you mind sharing how you found out it was posted here today?

I'm a regular HN reader, just happened to pull up the homepage and see it.

is brett glass of http://bikeshed.org fame still a wireless isp mucky muck?

I see this post come around every once in a while, and while the idea seems alluring, the fact is if you’re not in the right location and market, this kind of business will be unsustainable.

Maybe you can start your own ISP, but do not assume it will be in the same place you live. Analyze the country and find the pockets where you can thrive and grow a successful business, otherwise don’t even bother. Also bring a decent amount of cash and a good credit line.

Yeah, the story is usually like this:

* People found an ISP $NEWCOMER in an area where $ESTABLISHED has bad service and high fees.

* The newcomer invests loads of money into cabling the neighbourhood with fiber (or wireless internet) and offers deals with lower fees and better connectivity than $ESTABLISHED

* In a static world everyone would switch to $NEWCOMER and they'd make a lot of money, being able to expand to other places. But executives at $ESTABLISHED aren't stupid. They invest as well and lower their prices, too, just enough to be lower than $NEWCOMER.

* $ESTABLISHED waits until $NEWCOMER runs out of money. They rise prices again. It doesn't even matter if its price reductions mean $ESTABLISHED loses money in that region. The regions where there is no newcomer bring in enough cash to pay for the few regions with (temporary) losses.

Of course even if $NEWCOMER runs out of money, someone will buy their infrastructure and offer similar services, but there are more components in the competitive moats of established ISPs like (local) government granted monopolies in exchange for providing public buildings like school with free internet, or the bundling of streaming services with internet deals. Those moats usually are wide enough that there is no avail for small ISPs to break through. That being said, if you break through the moat, it's an insanely profitable business. DTAG had to invest tons of money until they got there and now they are in the place where the big ISPs can't harm them any more and the american daughter company is now the favourite of the DTAG shareholders.

A WISP, at scale (50 users or more), should have a cost per user of about $10/m. You can oversell bandwidth considerably, more than most people realize. I've specifically seen an 80 customer wisp reporting their peak time is only 230mbps. This largely makes sense as you people streaming video is going to be your biggest usage, and that really only average 15mbps at the worse. A $1/mbit for ISP bandwidth should be doable in most areas, and since it's wireless your costs per user are actually rather low, given a good tower.

But WISPs do make the most sense in areas where there is no good alternative (only DSL or Dish).

Yeah, the big problem is labor. if you are only making a few bucks a user, making enough money to make up for your opportunity cost (vs. working as a tech for a large company) requires a lot of users.

I've got a lot of experience serving the low-margin market, both for ISPs, WISPs, but mostly on the hosting side, (and hosting is actually easier; the less you have the customer on the equipment, the less time you spend fixing stuff they broke.) and let me tell you, low margin does not at all line up with low support expectations.

Even if you can charge your customer twice what it's costing you and making ten bucks a month per user, one tech support issue can end up eating all the margin that customer would have given you for years.

I'm certainly not saying to price it so you only make $10/m per user. I'm simply stating what your base cost is. $50/m should be expected with a WISP, as I tend to see rates of 50mbps+ from well run WISPs. I was just pointing out that it's hard to price someone out of market without dropping prices to ridiculous rates.

I'd love to know where you can get a fiber circuit for $500/m. (And even if you can, then you probably can't compete with whoever's selling you that circuit anyway.) When we started we haggled considerably.. (had to sign a 20 year deal) and managed to get a 2G/10G circuit for $1,500/m. (A 1G circuit would have been $1200/m)

Some places are cheaper than others, looking at this thread especially outside US.

In Poland you can get 1Gbps connection for 200USD/month and 5Gbps for ~600USD/m

These are not fiber circuits, ie dedicated bandwidth, they are shared residential services.

Those are not shared residential services, but normal business/ISP grade connection with SLA and so on.

I guarantee it's not dedicated. it may go over a dedicated line at some point, but I'm in the layer2 business running GMPLS and shared switched traffic and you're either on GPON or some Metro Ethernet gear. you are sharing the neighborhood bandwidth on a metro switch, or you're sharing both light spectrum (GPON) AND the neighborhood bandwidth. you could have 100g but it's still shared.

Do you think there's a company advertising 5gbps at $600/m for residential services? Those are dedicated fiber prices, all be it very good ones (I'd question the peering outside of Europe).

There's free.fr that sells 10G for 50EUR/month.

2G/10G? Normally I see it considerably cheaper for download bandwidth vs upload bandwidth, as your provider will just sell/reserve the upload bandwidth for servers, since servers use vastly more upload.

Of course, the rate you gave holds with the $1/mb I was assuming. You oversell bandwidth obviously.

It's probably 2G sustained, 10G burst of some sort.

>> You can oversell bandwidth considerably

Not as much as Comcast/$ESTABLISHED company can. The people most likely to switch to a WISP or alternative ISP are the people most likely to crush your bandwidth.

> A WISP, at scale (50 users or more), should have a cost per user of about $10/m. You can oversell bandwidth considerably, more than most people realize.

I run a small non-profit WISP (with price per user about $15/m), and the cost of bandwidth is not a significant part of the budget (about 15 %). Biggest part of the budget (about 60 %) is hardware and work for development and repairs of the network.

> since it's wireless your costs per user are actually rather low, given a good tower.

I would say that wireless has pretty high cost per user. To have good coverage and not overloaded APs, you target ~5 users per AP and 1-3 APs per node.

I think there will be more demand once people realize there are other options. I saw a post on HN a few days about someone starting a small ISP in NYC. I suppose NYC market is saturated as it can get.

One interesting I ran across is how mountainous villages in Nepal has their Internet connections. I was in Annapurna trail last month, and in Manang village (which has about 20 guest houses), there is one big satellite connection and everyone shares from it. The owner charges the consumers a flat fee. From what I hear, and it could be exaggerated, they pay about $500 per year. I think a remote location, one uplink satellite, and a few wired subscribers can be a viable business!

Also since these would be WISPs (Wireless ISPs), it depends on topology as well (signals can't go through mountains, backhauls need to have line of sight to multiple customers, etc). It's a tricky balance of multiple variables.

Thank you, this is fantastic. I am working with a 100+ employee company doing highly technical engineering work in a very rural location, currently trying to run their business off of 10 MBPS. They are getting quotes exceeding $500k from most people they talk to just get any additional bandwidth to them (presumable fixed line fiber), which has resulted in this critical upgrade being put off for quite some time. This guide makes me think they could quite possibly build out their own backhaul system to a neighboring town for substantially less money. Should make my next conversation very interesting.

it’s probably not wise to build your own backhaul if you’re not in that business. don’t know the particulars of course but most likely DIY is not the way to go.

startyourownisp is not really about that.

now if you can do a microwave or other backhaul that’s great but it still sounds like something you should be able to get someone to provide for you.

It’s definitely not the preferred solution, but the article here is trying to make the case for running a whole ISP for a customer base that’s less than the employee count of this single site. They also already maintain their own roads and I believe some high voltage equipment so they are a bit of a special case when it comes to being able to maintain infrastructure.

That being said, what this guide really gives me is an alternative solution to present to a CEO and IT manager who think that it is an unsolvable problem. I still think this ends with a 3rd party doing everything, but for substantially less than $500k (before monthly service fees).

You don’t have to do it all yourself. The article shows that for about $55k in the first year you could start your own ISP. This leaves about $450k for hiring some people who do know about running an ISP to get it off the ground and it’d still be cheaper. Maybe you could even start getting customers locally and break even in a few years.

he already said he could get someone to provide to him...for $500k. Why would it be unwise to build this if "he's not in the business?" isn't that literally what the above site is for?


when you build out physical infrastructure you are guaranteed to have real world physical problems, on top of deep infra tech maintenance issues that you can’t just let slide.

nothing is easy at this level.

$500k is likely cheap (but would have to see details) the main problem is that you are spending it all at once instead of capitalizing it over 20 years like the ISP is.

over the same period of time it’s going to cost the same. but in all the intervening years fixing problems is going to take longer, things will be less reliable, and you have to invest in things that are not a core competency.

again, not enough details are presented so it’s hard to say, but in general DIY is a bad way to go here.

meh, i disagree. If volunteers can build out networks like these which work pretty well, then I gather paid IT folks can as well. Since it has pretty low cost to start with, it could be used as a backup system until reliability issues have been worked out, and then used as primary with a fallback to the existing functional but painfully slow service they have now. Good monitoring could go a long way to making it viable. The extra money saved could then be used to pay for technology that does contribute to the companies core competency.

like you said, without details, i think the current systems admin could make this decision better than either of us arm chairing quarterbacking it. I certainly think it's viable though, when compared to operating under the poor dsl solution they currently have.

I've helped many rural locations get online. If you want to email me your address, I can see if there are more reasonable options available to you than $500k.

Email in profile.

500k? This is crazy

20k/mile to bore on average, sometimes 30k/mi, that's not direct bury though. that's for conduit

My problem is that I want to do this with fiber, not Wireless, which is considerably more expensive.

If only regulatory capture wasn't a thing...

If you want help with getting the costs down for fiber builds, hit me up. Email in profile.

tbh, your profile is a little too cryptic. I'm not sure I sent to the right address...

How about building software for local mesh and municipal broadband networks instead? Why does software these days have to reside on servers 1000 miles away? Not only is this slower, but also more wasteful of energy.

To the downvoters: can you actually respond and explain why you are downvoting this sentiment?

Would be cool to see community-run non-profit ISPs.

Both my hometown and where I currently live have municipal broadband.



I think there are quite a few in France:


https://bgp.wtf/ (Warsaw, Poland)

It is interesting how phrase 'community-run ISP' is interpreted in different ways. I would understand it to mean ISP run by independent non-profit organization or consumer co-op, but definitely not ISP run by government/municipal organization. But many posts in this thread (and also in other discussions) use it for municipal ISP.

That's all over the place in the Czech Republic. Really the best way.

One of my first jobs while I was still in uni was working for a small ISP. It was really fun, we had many base stations and mostly provided wireless connectivity. However it's all okay when it's summer to go and deploy/fix antennas but when it was -20C outside and you have to climb on the roof and fix cables or mount a new antenna.. Well, at that point I understood that I need to change jobs :)

Here are some talks from the guy behind that site: https://www.youtube.com/watch?v=cq4mWTQZmnI&list=PL1MwlVJloJ...

These are very dense talks with tons of information you'll probably not need unless you're reeeally going to do it, but it's good to watch for an overview of all steps involved anyway even if you don't.

The talks are from Altheapalooza from https://althea.net/, but they can be applied to non-Althea ISPs also (although you should consider Althea as it's great despite the shitcoinery).

Ethereum is not a shitcoin, not really sure why you felt that dig was necessary.

>Right now even with a very tech-forward customer base 500 customers will only rarely spike above 900 Mbps (0.9Gbps). That means that even at peak times a customer could still come on and run a speed test and get 100Mbps.

Is this still true?

A problem I had years ago was, I was living in an apartment that didn't even have a phone line. So I used 3G (4G was there but just not working at all) because the house Wifi was useless, in fact in the whole area this was a problem. So I started getting interested in using public Wifi (Freifunk) with a big antenna as upstream, and also provide it for the house eventually. But since then I realized that even on a normal Linux computer, configuring/routing multiple Wifi interfaces is absolutely non-trivial.

Eventually I gave up, but I bet Mesh Wifi are seriously becoming a thing in dense cities. Also some off-the-shelve routers allow 4G-USB-sticks as backup options. Fully automatizing this would be so awesome!

Where are the numbers on revenue here or what point of scale shows profit. They show a spreadsheet with 10 users costing $2,800/mo to maintain and $24k upfront. How could that ever be profitable at a competitive price point?

Am I missing a link here or something?

If one has their own ISP, can they pirate stuff without being caught?

You'll receive emails from copyright owners of they detect torrenting. But that doesn't mean you have to act on it.

I just want someone on the inside at comcast so I can figure out how to get their gigabit pro package, they say I'm like 200ft too far from this magical node.

I always wanted to have an open access RADIUS server so that I can resell my unused Internet. I personally think this should be allowed.

The problem with this is that consumer Internet access is priced based on the assumption that it will be "bursty". Heavy users are subsidized by light users, both on a moment-to-moment basis and overall.

Reselling consumer-level Internet access to means you're taking away the users that would otherwise be subsidizing your service (because they're buying from you instead) while still expecting the local ISP to shoulder a large share of the costs for bringing your packets to their final destinations.

That issue goes away if you pay for a dedicated line (IE, explicitly contracted reserved bandwidth) to an IXP or place somewhere will sell you bulk transit, and you're paying for that too. That's fine. The problem is with expecting to make a profit by being the ISP to your whole block, based on a $100/mo GBPS fiber drop.

Not to say that the ISP mono/duopolists aren't loathesome--excessive market power is the root of... not all, but lots of socio-economic evil. The economics of telecommunications is complicated.

You mean resell the thing you agreed not to resell?

I’m starting a WISP in the Oslo area if hackers in that area want to join forces. I could use a native Norsk speaker!

I would like to join forces and bring a Norsk speaker with me. Please contact me. MerikAtFiberhoodDotNl.

This is from a broadband provider perspective but this company vetro is enabling people to do lay smart fiber. This could help someone lay do a fiber plan for their local municipality https://www.vetrofibermap.com

Just last week I was wishing exactly this existed. There are so many businesses where the main hurdle to get over is the basic institutional knowledge.

I wonder how cheaply it could be done for 1 customer (my parents house has terrible wireless internet and I'd like to move home and try doing a startup)

>I wonder how cheaply it could be done for 1 customer

Depends on your definition of cheap. Your biggest expenses would be monthly bill to your fiber provider, then any necessary rent for backhaul sites, then the up-front cost.

I can't say for sure the monthly cost but it would definitely be in the low four figures.

About how much money should you be prepared to spend if you were to follow this guide exactly?

There's a spreadsheet linked [1] that says (for 10 customers, approximately) $24,000 upfront and $2,800 monthly.


I found out the hardest part was finding the backbone or a provider. I am in the middle of deploying a community WISP but finding a provider besides Spectrum has been the biggest hurdle.

These days on some websites you can sometimes find "One Weird Trick to Cheaper Rural Broadband" ads by Althea. The phrasing is certainly interesting.


Does anyone have thoughts on the viability of starting your own WISP considering Musk has already done his first launch of satellites?

Get yer money pit here! Roll up...roll up!

This is awesome, thanks for writing. I have no interest in starting my own ISP but still read all of it!

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact