Download a PDF/EPUB version from
Creating Internet in the Apocalypse
-by Nate K.M. Bowie
Licensed Creative Commons
posted to Hive.Blog on 2/26/2022
I’ve been watching the Walking Dead lately, and even though I’m a bit
behind on the trend (as I often am) I’ve pretty much caught up to the
latest portion of it that’s out as well as most of fear the walking
dead. Anyway, I thought I might put out a little thought experiment,
what if you wanted to create your own internet within the apocalypse?
If so, what are your options and what could you do with it after
building it. Believe it or not, it would be possible to host a live
stream on an internet made from scratch in such a scenario (though that
won’t work with most of the options laid out), and there would be
legitimate uses for such a so called ‘internet’.
Before I begin I would just like to point out that this is mostly just
a thought experiment. Networking is a good chunk of my job, but at the
end of the day I haven't done things such as set up a GMRS mesh network
(I’ll explain that in a bit) so no guarantees that things would work
exactly as I lay them out. Regardless, time to get started in somewhat
of a disorganized fashion.
Ultimately, the internet is simply built on connections between
computers. If I load a file on my phone’s browser (using HTTP) from a
server via cellular data my phone uses radio waves to connect to the
cell tower with the requested information. The information is then sent
from the cell tower to the cell company’s servers, probably through a
high speed cable, where it’s then processed. If it was a domain that
was requested it looks up the DNS records to see which IP is associated
with the domain, then with the IP known (or if the IP was already
known) the connection is made. Say it’s a server in a building that has
internet provided by a cable internet. In that case that connection
goes from my cell phone company to the cable internet provider, then
from that cable internet provider to the server in that building via
the cable internet connection. The file is then sent back in ‘packets’,
which are just pieces of the file, which are received and assembled
into the file.
You might be familiar with my writings on IPFS, in which I spoke about
the differences between IPFS and HTTP. In this case, however, they’re
both connections between devices running on IPv4/6 so the above example
works about the same regardless. Long story short, as long as your
electronics have the capability of communicating and have been
configured to do so they’ll work. And even though the modern global
internet has a lot of complexity, if you’re scaling it down it gets a
lot simpler and is legitimately buildable.
I’ll start by explaining some of the important ‘layers’ that build up
such a network. I apologize if it’s too many random aspects to quickly,
but once I get past explaining all the different parts I’ll give some
examples of them being in use so even if I do a bad job at explaining
the parts individually they’ll start to make sense once they’re put
So, on to the plans. To start out, if you want to build a network you
need power. In the Walking Dead they seem to run into Schrodinger power
reserves: there is simultaneously unlimited power and no power, and
we’ll never know which is at play until the writers decide if it will
make a good plot point.
There are a couple of ways get power even if the grid is down. What you
intend to build within the network, as well as what supplies you have
on hand, dictate your options. Right off the bat, we can probably rule
out sources like geothermal, hydro power, coal, nuclear, most
windmills, and just about anything else that will require serious
service by experts. This mostly brings us to generators, solar, and
Generators are the simplest option, as long as there’s fuel you can run
them and have the exact amount of power required. There’s a few caveats
to this, however, the most obvious one being the fact you need fuel.
Cheap home generators can run off gasoline or propane, and there are
diesel generators as well, but obviously there’s a limited supply of
fuel. Gasoline goes bad, though fuel stabilizer is cheap, but unless
you’re growing crops for fuel ethanol or oil for a diesel generator you
can’t get more fuel once it’s gone (and good luck making propane).
Beyond that, even if you were using ethanol or plant based oil for
gasoline or diesel generators respectively it’s my understanding that
engines wear out pretty fast if you’re using those.
Furthermore, a lot of generators won’t be safe to use with electronics
out of the box, at least not the kind you would pick up at a hardware
store, so you’d have to take that into account as well.
Your next option, which seems obvious, is solar panels. They come in
all shapes and sizes, from huge with high outputs to ones small enough
to fit in a backpack, and are a go too for all sorts of people who want
or need power outside of the electrical grid. The problem comes in
though when you need power and the sun is not out. While solar panels
are probably the best option here, depending on your setup you might
need a beefy power storage system.
Last in terms of power, there are some options for you to generate
power manually. This usually includes thinks like wind up radios that
will charge via USB or chargers that will generate power while you’re
riding your bike. They certainly won’t generate enough power for most
things, but could probably charge a mobile device, radio, or
flashlight, so it’s worth mentioning.
Once you have power though, there will be two limitations that you’d
have to contend with. First, power will pretty much have to be used
where you generate it. You can run devices or charge batteries, but
otherwise you’re limited to about how far you can run an electrical
cord from your source of power.
Speaking of charging batteries, that’s another thing you might run in
to. If you’re using something like solar then once the sun goes down
you’re limited solely to what you have in storage. Batteries like car
batteries are not meant to have a constant pull on them, and will
probably fail pretty quick if you try to run a bunch of devices off
them day in and day out. Something like a deep cycle lead acid battery
intended for boat engines might work decently, but if you were able to
get your hands on a number lithium-ion batteries that would be ideal.
There are also battery systems often known as solar generators that
come pre-built for power storage.
Additionally, many devices from phones and laptops to portable radios
have built in rechargeable batteries, and some are power efficient
enough to be recharged with a USB power bank. Depending on the scale of
what you wanted to build, they might be your best bet for reliability.
Okay, so you have power, now you need devices to be able to
communicate. How ambitious you are depends on what kind of lines
(metaphorical or literal) that you need to lay out. Just keep in mind
some types of hardware might be incompatible with some of these options.
Speaking of metaphorical and literal lines, your first option is to
literally lay out cables. Stepping outside and laying out or burying
500 feet of weatherproof Ethernet cable is actually not actually that
hard. Cables are also a pretty efficient means of getting data from one
place to another, but unless you’ve got the equipment and manpower to
lay cables on a massive scale there’s a limit on the amount of distance
you can cover with them.
WiFi & LANs
(I'll use WiFi
and LAN interchangeably here)
The home WiFi network is pretty standard now: even if it’s not
connected to the modern internet it’s still very functional as a LAN,
and if you’re building a mini makeshift internet then it can provide
access to that as well. Furthermore, you can get the benefit that most
modern devices can connect to it without any hassle, many modern
devices can create their own network, and something like a router or
laptop can create a network that can bridge a connection from something
like Ethernet to mobile devices by creating a WiFi network. You can
also find routers capable of projecting WiFi five to ten thousand
square feet, with extenders that can increase that indefinitely, so you
can provide a pretty big area with connectivity.
WiFi and LANs also have the benefit of getting data to a large number
of devices easily. Sure, cables are a really effective means of getting
a high speed and low maintenance connection between a couple of
important devices, but if you start building out a huge network with
only cables then you run into issues where you might need to start
including switches (which require power themselves) and a bunch of
complex configuration issues. Even if you rely on cables a lot for
building up the backbone of such a hypothetical network in the
apocalypse you’d still be best left of using wireless communication at
the edges of that network if you plan on having plenty of people
connect to the network to utilize it with their own devices.
Long Range Radio
Beyond WiFi, which uses radio signals at fairly short ranges, you can
actually get data transferred pretty far with radio waves. Keep in mind
that your cell phone is pretty much a handheld radio with a computer
built in, and satellite’s are radios with big reflective dishes to send
and receive data in a precise manner.
One option for longer distance radio is to use a device that functions
similar to a proxyham device. If you were unaware, a proxyham was a
device that was originally developed to access WiFi at relatively long
ranges (supposedly up to nine miles under ideal conditions). While the
proxyham was never actually released to the market, it’s possible to
build a makeshift device with a USB WiFi receiver, a flexible USB
cable, and a satellite dish or even a metal colander.
Next, there’s the ability to use analog radio systems to send data.
Products like the Gotenna Mesh allow you to create a mesh network using
GMRS band radio signals, allowing the sharing of data/communications.
Similarly, some Ham radio operators use APRS, which allows packets of
data to be sent and received via radio signals. There are also FRS/GMRS
radios which will allow you to pair them with your phone and send and
receive what amounts to texts between users over reasonable distances.
Now, when it comes to radio signals
in real life as it stands now there are certain limitations to what
you’re allowed to send, how you do it, and what licenses you’d need;
though I guess if this is a thought experiment about networks in the
zombie apocalypse then you can get pretty damn creative using radio
signals to transmit data.
Despite all these complicated options it’s still possible to deal with
distance by simply walking or verbally transmitting data with a radio.
Obviously there’s a limit to what you can verbally transmit over a
radio, but it doesn’t require much in terms of setup. You can also
simply bring a device from one place to another and connect to an FTP
server or transmit Briar messages that you traveled to bring (more on
that in the software section), or just bring data on a thumb drive/SD
card/hard drive from one place to another.
Once you have power and a means for devices to communicate, the next
logical step is address the devices that make up this hypothetical
network. You have a lot of options, and of course a wide variety of
devices can connect to one another and work in tandem, but I’m mainly
going to divide this up into three groups of devices.
Servers, Desktops, and Laptops
Okay, so you might be wondering why I lumped budget laptops and state
of the art servers in the mix, but a lot of this comes down the
software they are capable of running and not their specs. Running a
full blown computer of some form will allow for much greater processing
power, and also allow you to do more advanced things such as
configuring a DNS setup to use real domains in this hypothetical
makeshift setup. If you were going to store a lot of data, run anything
complex, or have a very complex network with a lot of connectivity than
a computer of some sort would be the way to go.
Breaking things down a bit more, dedicated servers (and many powerful
desktops) have much more power than laptops and other lower end
devices. If you were going to run things like video streaming, or
wanted to have a complex library of data, or had a large sprawling
network then something with a bit more power would be vital. This comes
at the cost however of requiring increased power and possibly increased
complexity. Not to mention you have to find and prepare such a device.
Laptops on the other hand require much less power and usually have a
battery system built into them. They also have the bonus of usually
being able to create their own WiFi network and utilize Bluetooth,
while still having the ability to run software like Windows Servers and
Linux distros; and have ports such as USB and Ethernet. While they lack
in computing power, if you’re building a network like this you might
not need a ton of computing power unless you were really building it
big, so they’d probably suffice in a lot of circumstances. They are
also really common, and you can get by a password super easily –
unencrypted devices can be bypassed with ease, and even if the hard
drive is encrypted you can still wipe it and use the laptop itself.
Mobile Devices & Micro Devices
Since this is a theoretical network for the apocalypse, however, we can
still go a lot smaller than laptops, and the first of that is phones.
Phones are pretty much mini computers these days, and can easily run an
FTP server, create a LAN network, and host other low power processes.
They are also very power efficient, and would allow you to pretty much
run a network from a $20 solar charger. They are of course unable to
accept most wired connections and physical storage, however, and lack
the ability to run certain software (so don’t expect to run DNS from a
Phones and tablets would make a great end point. For example, you could
have something like a raspberry pi receiving packets from a radio, then
connecting to a LAN created by a phone, and then anybody who connected
a device to that network could be patched in to the larger network on
the other side of that radio. On the smaller side of things, a phone’s
LAN could be used as the only device to device communication, with a
phone being located in one main building and every time somebody wished
to share a file or leave a digital message they would go to the
building and connect to the network. Last, Android would probably beat
Apple simply because it’s easy to side load on Android, but if the
Apple store goes down then you’re stuck with whatever you already
installed. That said, it’s not like Apple would be useless since you
could still try to jailbreak it or simply use already installed
The last component I thought I would cover is micro devices like a
raspberry pi. While a device like that is probably less powerful then a
mobile device, it’s also geared more towards networking so it’s sort of
equal in it’s ability to perform small tasks. A raspberry pi could be
connected to the network through a LAN or wired connection and host
lightweight content or perform tasks such interpreting radio
transmissions to allow devices over long ranges to communicate.
End User Devices
Last, I might as well touch on end user devices as well. An end user
device here is simply any device serving people instead of the network.
You might have a number of devices hosting content, and a number more
making sure communication between the devices hosting content works,
but an end user device is something that a person uses for themselves.
This could be a computer that is plugged into the network or connects
to a LAN, or a mobile devices that connects to a LAN, but it’s so a
person can perform a function such as communication or file sharing
(probably the same things you use your devices for now). These devices
don’t matter too much as it’s up to the device’s owner and not the
network to keep them powered and running smoothly, and the also have
the benefit of not causing disruptions when they’re turned off or
That said, phones and tablets connected via LAN is the most likely end
user device to be found on a network like this. Devices that make up
the network itself may need more capabilities than a mobile device, but
in order to do most functions like chat or share files a mobile device
is ideal given it’s built in battery, portability, and minimal power
Power, connections, and hardware make a network possible, but it’s the
software that gives the functionality to it. Now, given this is all a
hypothetical thought experiment, I could just say that you’d need to
program everything from scratch yourself, but realistically getting
modern existing software is the way to go here. Obviously lots of
modern software like Google Chat is useless if Google is down; and
protocols like IPFS, while great, given their hash identification
system and the fact that this network would likely be too small and
inconsistent to allow decentralized tools like that to shine. Instead,
there’s specific tools however that will fill those roles nicely, many
of which you might have found in a home network a decade back. Also,
because this thought experiment takes place in the apocalypse, security
is less of a concern so no worries about outdated or somewhat insecure
First off, you’ll need operating systems and general tools like that.
Windows and Linux installers are certainly crucial, a portable Linux
distro like Puppy Linux would be ideal for bypassing any password
protected computers you come across, and software to provide DNS or
packets via Ham/GMRS radio and plenty of other networking software
would be crucial if you intended to use them.
Additionally, you’d need generic utility software. Obviously if it
needs to check for a subscription then the software is useless, since
even though you could change where the domain points via DNS
configuration you would still have to spoof a certificate to sign the
software so things like MS Office and Adobe software is out of the
question. Instead, document editors, photo editors, note taking apps,
calculators, calendars, file compressors, ebook readers, music/video
players, audio recorders, html document editors, and plenty more would
be useful to have options for on multiple operating systems for
computers and mobile. Also, while in an apocalypse GPS would be useless
within a couple weeks, offline GPS would be useful at first and mapping
tools like OsmAnd would still be useful for understanding the terrain
and making markers even if GPS no longer works.
Next, communication tools would be extremely useful. Most messengers
are useless without a central host (Apple Messenger, Google Chat,
Signal, Telegram, WhatsApp, etc), but there are still a couple of
options. Email could be configured via DNS, but email would probably
bit a bit clunky to configure and unnecessary for a small network like
this. Federated messengers like Matrix or XAMPP would be a bit simpler
to use, and there are also messengers like fire chat which work off
Bluetooth only, though I have not used fire chat personally and
federated messengers would still be a bit finicky. That said, in my
opinion, Briar should be the go to of any apocalyptic network engineer.
Briar operates in mesh network style, providing all the data one device
has to another whenever they can reach each other, meaning it would
work perfectly in a situation when devices cannot always reliably
communicate with each other. It is also completely decentralized (not
relying on any centralization or federation), it can communicate though
the internet, Bluetooth, or LAN, and it can also host group chats,
private chats, and discussion boards.
Next, you would definitely need a way to share files. These files could
be videos, photos, maps, ebooks, notes, lists, or whatever else, but
the ability to replicate files and share them with others would give
anybody an edge in the apocalypse. Both an FTP server, which could be
run of nearly any device, or a Samba server on a Linux computer could
fill this role nearly perfectly. Additionally, there are also plenty of
programs and apps that would allow somebody to automatically sync files
Continuing on, depending on your setup if you wanted to host web sites
on this hypothetical network you would probably want to look into web
scripts. Web scripts allow you to create a complex website with no
coding, such as WordPress blogs or discussion boards. That said,
complex websites take up more resources, and a discussion board you can
load with a domain name in your browser might seem fancy but something
like Briar would work almost as good at a tiny fraction of the power
cost and setup complexity.
Ending my examples, you may also need specific software for any tools
you are using. Want to create a GotennaMesh network? Then you’d
probably need to get software from GotennaMesh. Same goes for Ham radio
packets, or phone jailbreaking software, or equipment repair software,
or anything else specific to your setup. Beyond that, there’s probably
a million other examples I could give, but I’m lazy and these items are
what came to mind.
Once you have a network setup the next question is how long and often
to keep it online. The simplest solution would just to leave it on
24/7, but that may or may not be the best solution. If you have the
power then certainly having communication capabilities at all hours
would be ideal, but when budgeting power there might not be enough for
it to be worth it. It may be found that leaving it online 8-12 hours a
day can allow everybody to use it when they need it, but also cuts back
on power usage so you know you won’t run out. Or, maybe have multiple
different settlements each with their local networks all bridged
together through radio communications, so you leave some or all the
local networks online 24/7 but only run the radio communications for a
few hours each day to cut back on power.
But again, given this is a hypothetical thought experiment, there’s no
real answer to this. A phone might take a week to burn through the same
power as server does in an hour, and you may have a ton of power to
burn or you may be so limited you can’t even keep that phone on 24/7.
The first, and biggest by far, reason why networking like this would be
useful in an apocalypse would be instant communication across long
distances. Most of this communication would ideally be done in text
form, making it silent if you need to alert people for strategic
purposes without making noise, it is super small in the amount of data
it takes up, and can easily be read and stored so you’re not limited
solely to what somebody is saying in real time. This can be broken into
First, as the old saying goes, knowledge is power. If you can send a
message to somebody miles away about something time sensitive that can
be extremely valuable. From getting help to somebody who’s injured, to
warning of an attack by people, zombies, or whatever other apocalyptic
threat that would fit this thought experiment – in addition to real
time collaboration while responding to said issues.
The second, a bit more mundane, is the administrative side of
surviving. Being able to share files on a portion of the setups
discussed it would make it possible to share information on anything
from people, to resources, to collected data. In the apocalypse that
accompanies this thought experiment it would probably be extremely
useful to share things like spreadsheets of resources that are needed
and that are in surplus, information on people and who is good or bad,
weather, and any other similar form of data.
But file sharing isn’t just limited to spread sheets. Ebooks, guides,
videos, audio books, software, photos, and just about any other file
could be shared through a number of different setups discussed. All
these different files could be huge in providing information to people,
and as said before, knowledge is power. It could also be used to share
entertainment, which would probably be a much needed moral boost for
any apocalypse survivors.
Last, a network like this could preserve data. Everything from game of
thrones to scientific research is, to different extents, an
achievement. While sticking copies of data on a thumb drive might be
destroyed by bitrot within a few years to a decade, as long as a
network like this was up and running it could be preserved for a lot
longer. Now of course if factories are going to stop producing
electronics forever then eventually the data would be lost anyway, but
a network like this could probably be kept running for a good 20-30
years before the worlds mono-crystal solar panels start dying out.
Okay, so finally here all that explaining can pay off and we can make
some examples example. Lets say we have one big town, a hamlet near
that big town, three small towns, an isolated building, and a small
house. There is networking hardware specific to each location, each
location has a local WiFi/LAN network, and they transport data from the
local network to other local networks using radio waves for long
distance networking. Also, let’s assume everybody is using the Briar
app as their primary means of communication.
Starting with the big town, lets say they’ve got lots of people and a
fair bit of power. Their local networking hardware includes a desktop
PC, a router, a small server, and a radio that all stay on 24/7. The
radio is connected to the desktop PC, and the desktop PC and small
server are connected to the router which is very powerful. The PC has
an FTP server on it, allowing anybody to upload or download files from
it, and it also allows handles packets from the radio. The server has
DNS setup on it and also hosts a couple of really simple websites.
The router provides a WiFi/LAN network that covered most of the town.
Given the PC and small server are connected to it people can access all
content on those devices (e.g. FTP file sharing & websites), are
capable of peer to peer data so they can communicate through Briar, and
thanks to the radio anybody on the big town network can access other
networks if they also have a radio set up to send and receive internet
The hamlet is just outside the big town, so they have a pretty unique
setup. They have a proxy ham type setup that is capable of connecting
to the big town’s LAN network from a half a mile away. The proxy ham
like setup then creates it’s own WiFi network, but connecting to the
network in the hamlet is as if you connected to the network in the big
town. Despite the fact the hamlet doesn’t have a radio setup or any
local devices providing them with content, since the hamlet’s wireless
network is just an extension of the big town’s network they can piggy
back off big town’s hardware and use the servers and radio setup just
as if they were close by.
Town one is a fair bit smaller so they have a little less in terms of
hardware. They have a radio, a laptop with an FTP server on it that
also handles radio packets, and a WiFi router connected to the laptop
that covers the center of the town with a LAN. People connected to the
LAN can share files through the FTP server and communicate with others
on the network through Briar. The radio antenna allows them to connect
to other town’s servers such as big town and communicate with other
LANs that also have a radio antenna, though the connection is fairly
slow over radio when compared to a LAN network. Town one lacks as much
power, however, so they only leave the laptop, router, and radio on for
12 hours a day, then turn it off to preserve power unless they need it
on for one reason or another.
Town two also lacks large amounts of power, but they go about this in a
different way. Their network consists of a radio, a cell phone, and a
raspberry pi that is left on at all hours. The raspberry pi is
connected to the radio and allows connection to other systems linked up
through the radio. The raspberry pi is connected to a LAN created by
the cell phone, and the cell phone also can run tools such as an FTP
server; though the trade off for such lower power requirements is the
limited space and computing abilities of the raspberry pi and cell
phone. Nevertheless, devices can still be connected to the LAN created
by the phone and operate it like any of the other LANs on the network.
Town three is limited in their power and the smallest of the towns, so
all town three has is a laptop that is connected to the radio. Though
the laptop could create a network, instead it is usually off, and the
laptop and radio are only turned on when they are used with the laptop
being the only main gateway into the other LANs that make up the
network. When turned on Briar messages and the like can come in, and it
is capable of accessing data on LANs thanks to the radio, but then it
can be turned off to limit the power consumption and complexity of the
local town’s hardware.
Lets say that this isolated building is an apartment building. This
building, thanks to it’s concrete walls and the hills between it and
the other towns cause it to be unfeasible to use radio to communicate
with other towns. While a repeater might be able to be set up on a
hilltop and the roof of the building, it’s not within the means of the
In this case, they have a local laptop that hosts a Samba file server,
which operates similar to an FTP server. This laptop is situated on one
side of the building and is also creating a LAN network. On the edge of
LAN range there is a WiFi extender that provides a boost to the LAN’s
reach and covers a good part of the building. With file sharing on the
laptop, and messaging through Briar, the people in the building can
communicate with each other. Additionally, wherever somebody travels
from a town with a radio to the building that is also using Briar they
can connect to the network in this building and automatically
distribute messages to this network, as well as upload whatever files
they wish to the Samba server.
Last, we can cover the random house. Lets say this random house is home
to a family who farms plants and livestock, as well as scavenges. The
only hardware they have there is a phone, but whenever somebody drops
by from another town they can connect phones through Bluetooth or a LAN
temporarily created by the phones, the have all messages coming from
and going to the larger networks exchanged. They can also exchange
files and the like whenever wanted. This would allow them to organize
things like trade that utilizes the larger set of individual networks
without having any hardware setup at that random house except a phone.
Further, since they need very little power they can suffice with a
super tiny solar panel or with manual power generation options like a
hand crank radio with USB output.
Well that’s about that. I hope you found this interesting, it was
really intended as more of an interesting learning adjacent piece than
anything resembling a tutorial. Stay safe, have a good one, and if
you’re really interested in any of the concepts here you can always
give them a try in the real world.
All content in /tutorials is
licensed Creative Commons
unless otherwise specified. Feel free to share it around and re-use it,
but please give me credit if you do so (and would love if you sent me a
message or something). You can also
send me a tip if you found this helpful, but no
pressure - this is a hobby.