The Internet during world war
oecd-forum.orgThe policy suggestions are fine but the reasons are kind of silly.
I've never been in a conflict zone but I'd assume securing food and water is #1. Reliable electricity access might be a distant nice to have and cell phone or wifi signal? I'm guessing that's a no.
Decentralize, increase redundancy, eliminate single points of failure, sure all good.
But if we're really talking about internet like communication during wars then I would assume what we need to talk about is off-grid power generation, packet radio infrastructure build outs, hardened underground storage containers for computing resources, etc. Without those you aren't going to have a phone that turns on or connects to a tower.
Maybe this is a marketing play to sell things we should be doing (like community networks) to people who otherwise wouldn't listen. If that's the approach than best of luck
It is not feasible to either fight a war or do basic things like ship food without communications.
During World War II, to coordinate the UK and United States war efforts, at the peak several thousands of teletype channels, were in continuous operation across the Atlantic (something like ~10 kilobytes of text per second) and priority mail shipments by plane (often shrunk to microfiche to reduce weight) were measured in the tonnes per week. The Allies even spent around a billion dollars (inflation-adjusted) to create an implausibly-complicated system [1] to allow encrypted voice communication between high officials over shortwave. Allowing FDR and Churchill to speak real-time, even just for a few minutes a week, was considered just that important. And back then, they were used to doing things with much less coordination from afar.
> The SIGSALY terminal was massive, consisting of 40 racks of equipment. It weighed over 50 tons, and used about 30 kW of power, necessitating an air-conditioned room to hold it. Too big and cumbersome for general use, it was only used for the highest level of voice communications.
And now I can whip something similar in one afternoon, and run it on my phone. Nuts.
I don't think GP was downplaying the importance of communication for waging a war, I think they were saying civilian internet access would not be feasible during a war.
It was more about whether these legitimate concerns deserve to be columned under war planning.
I know military science is its own field with many long running scholarly journals, precisely 0 of which I've read. I fully acknowledge I speak from a point of ignorance, I'm just surprised and would like to know more
Your first assumption is wrong. Food and water are necessities, of course, but you can survive without either for days (outside of extreme weather environments) and they're also very portable. Communications are an urgent and ongoing need, for military actors and civilians. Without them you are not only in a bad situation, you have little idea of where to go. Military communications manuals contemplate up to 5 levels of communication for different contexts, from wraparound digital services to smoke signals and the use of secret symbols for identification/navigation.
You are right about communications infrastructure, which is a favored target in any militarized conflict. Mesh networking offers a workaround to communications breakdown, albeit at the cost of bandwidth and deployment effort. LoraWAN devices are quite cheap at the margin (under $10 in bulk) but hardening, powering, and installing them is labor-intensive, to put it mildly.
What the war in Ukraine and protests in China have showed the world is, democracy in desperate times survive on Wi-Fi, AirDrop’d screenshots, and dozens of single port phone chargers hanging off power strips.
They don’t even bother to use multi-port chargers, let alone DC to DC converters, or cars rigged for power generation. Not to mention open-source encrypted software-defined self healing mesh network microwave radio protocols.
Maybe wartime is not the best to introduce alternative infrastructure that are inherently new; maybe only what’s there peacetime works.
And if that is the case, what we would need might be such items as, sideloadable resilient mesh apps for teenagers, appliance-fied server for game sessions in University dorms, and parallel chargers for camping.
You assume the presence of electric power.
My acquaintances in Ukraine report using various batteries, and converting UPS devices to run on larger batteries, car batteries, etc. Electric power is now intermittent there, even in the capital, because Russian strikes are especially targeted to destroy the electric infrastructure.
I think GP was referring to the photos that circulated of liberated Kherson with rows and rows of fully populated power strips charging peoples' phones as many of them use them. This was prior to the concentrated energy infrastructure strikes.
> democracy in desperate times survive on Wi-Fi, AirDrop’d screenshots, and dozens of single port phone chargers
Democracy is utterly dependant on a well informed population.
If the fair and balanced media fails (or access to it fails), then democracy is in peril.
It's an interesting thought experiment! What would happen if electricity ceased suddenly.
If there was no electricity where I am at, the apartment would soon start to cool down. It is right now something like -6°C outside. I think it would take some days to cool down to that level.
Electricity keeps my food refrigerated and frozen, both at home and in the shop. Well, I should put the frozen food out on the balcony (and cover them up so that they don't melt if the sun shines during the day).
Kitchen oven and stove, and water kettle run on electricity. There would be very little cooking going on! Not to mention that electricity is needed to pump water from the ground level up to the apartment. So I should to fill all available containers with snow and take them inside to melt (before the apartment temperature drops below freezing).
Seems like loss of electricity would mean loss of heat, water and food in rapid order. It's difficult to imagine this loss to be very localized, so my neighbours would likely be suffering the same, perhaps the whole suburb. Without apartment or street lights less reliable members of the community might try to take advantage of the situation. Perhaps not, it's not a particularly restless environment. More resourceful members might come up with solutions to problems, but I can imagine the situation to be quite chaotic.
Without communication channels I can't ask if a friend or family member can support me, or if they need support. Or indeed if it is safe to move, and if so, how and where. Lack of street and even apartment lights might be a challenge to many and make traffic overall quite a bit more dangerous.
I don't know how reliable electricity would need to be. But some reasonable level of access to electricity is definitely more important than nice to have.
> If there was no electricity where I am at, the apartment would soon start to cool down. It is right now something like -6°C outside. I think it would take some days to cool down to that level.
I mean, I lived this during finals week in college. During the day solar heating helps some. At night, with good insulation in the home it'll take some time to match outdoor temps, and a winter rated sleeping bag helps at night. One of my idiot roommates was using a sleeping bag as a comforter and complained at how cold it was on night 1. After I pointed out what the zipper was for, he reported it was toasty warm on night 2. You can do the same thing without a zipper, its just slightly less effective.
Also helps: the water heater runs on gas. So you can take a warm shower. In the dark, since most bathrooms have no windows, just exhaust fans. But the furnace, while it needs gas to heat, required electricity to move air. I'm not sure why, but water mains don't usually freeze despite the usually cold winters (though indoor plumbing requires certain precautions in the cold).
Cellphone towers usually have some batteries and backup plans in these situations, which was less useful back then before data plans existed, but meant you could still reach friends in case their power was on.
Finally: emergency generators are a thing. So as long as it isn't so cold that diesel freezes, emergency services will be available to the community in a school or something. In cold climates there are enough of these that in the lead up to the ice storm that caused the mess there were PSA campaigns to get them tested and vetted ahead of time -- if you just energize the lines without installing a cutoff switch, it can kill line workers trying to repair.
power outage is a common condition of modern institutional planning exercises.
You seem to be assuming "there is a world war" means "everywhere is a full scale conflict zone", which it doesn't. In reality, the severity of fighting will vary significantly. Some areas may face challenges with food / water, other areas won't. While the power grid may not be reliable in some places, generators, solar power and local power generation will mean that most people will have some level of access to power.
One of the points the article is making is that stuff like "off-grid power generation, packet radio infrastructure build outs, hardened underground storage containers for computing resources, etc" will be important in areas with active fighing / bombardment, many other areas may face entirely different types of challenges.
Do we have a process for quickly revoking CA certs owned by enemy countries? Do we have locally hosted mirrors for important content / services / tools that are hosted across undersea cables? The internet has been more and more integrated into our work lives and it does bear some thought to make those systems more resilient so the economy can maintain some level of functionality in a world war.
> I've never been in a conflict zone but I'd assume securing food and water is #1. Reliable electricity access might be a distant nice
This is very wrong, everything NATO/ISAF related excepting the absolutely smallest temporary COP and FOB in Afghanistan had at minimum one basic VSAT terminal with IP link to the outside world.
You can't have modern C4I systems without data links. Note that a lot of what was implemented was not actually "The internet" as we know it, though there was lots of commercial DIA, there's plenty of ways to use two-way satellite capacity for entirely private networks.
The above was a firm rule and ground truth even 15-17 years ago, in the latter stages of the conflict (before the US lost political will to continue in 2021, Thanks, Trump and Pompeo and Biden...) data links for anything military related were even more crucial.
What you’re talking about is worst case support, which while important, is not typically required all the time.
Portable battery packs and generators, or power on sometimes, but not all the time would be normal.
Same with internet/ISP connections.
Cellphone towers and WiFi would usually work, unless there is a big offensive going on at the time, etc.
That means a lot of the less severe stuff is worth talking about too.
Needs to be looked at from all angles including:
+ Shared physical infra Trans-oceanic cables being the most fragile. However, the risk here would not immediately be of the power-food-shelter variety. Satellite-based comms will potentially add resiliency. Intra-region (depending on the region), there is less fragility (more independent fibers), and emergency tactics such as P2P mesh networks aided by swarms of drones, balloons, satellites becomes interesting.
+ Shared software infra Because my network talks to your network, and/or leverages common (or at least sometimes cascading) structures like DNS, NTP, shared BGP routing tables, what risks do I have? How could I mitigate them proactively? What would I do in an emergency which wouldn't have worse consequences (trade-offs are fine)? This category of risks can be difficult in that the events might not be obvious at first (unlike most physical infra events), but that could make it worse...
Sort of a related tangent to the "the internet was designed by ARPA to survive nuclear war" trope that is trotted out occasionally...
It's NOT that the Internet specifically was designed or intended for this, it was a research network. The pre-existing AT&T Long Lines network which carried telephone traffic around the US48 states also carried the DoD's AUTOVON network data links for their own phone system, and data links between things like SAGE direction centers.
Many of the cold war era hardened AT&T Long Lines bunkers, underground sites, special buried mountaintop sites and such all pre-date the earliest days of the ARPANET.
There's mountain top long lines sites out there now that couldn't be duplicated for less than $50 million. Money was thrown at this in quite a profligate manner - same as money was spent on various early generations of ICBMs, strategic air command 24x7x365 standby and patrols, the DEW line, and such.
In the era before inter-city singlemode fiber optic cables were a real operational reality these places were absolutely crucial.
The data links between sites in many cases rode on top of these networks in the earliest days of IP.
Fascinating. I think that the capacity of nuclear weapons and MAD decreased the likelihood of future world wars to so low as to be effectively none.
But this article kinda opens the possibility of world war taking place entirely in the data/communications realm.
I think future World Wars are definitely possible with a few caveats:
1. There can be nuclear combatants but the fighting is primarily on the soil of non-nuclear countries.
2. Because of #1, the war more like World War 1 than in World War 2: some sort of peace treaty that doesn't stipulate unconditional surrender. There is no regime change nor post-war occupation of nuclear powers.
3. There is an unprecedently high level of communication between the nuclear combatants as both sides try to reassure the other that they aren't going to unilaterally launch nukes (there may be nuclear brinksmanship in public though as propaganda tool. Backchannels will be all about nuclear deescalation)
> no regime change
I suppose a regime change can come from within, as a consequence of a lost war and economic hardships. I suppose Russia has a fair chance to experience just that, once it loses.
Valid points, although risks seem to depend on the scale. A worldwide conflict would have different risks than two-state conflict. I don't see any reflection in the article or linked materials on the lessons from Ukraine, and the potential role of satellite internet in bridging localized infrastructure gaps.
I don't think many lessons about a potential world war could be gleaned from Ukraine. A world war is going to necessarily pull in all of the major powers which means that there won't be a single front, navy and air force will be involved and strikes deep into territory on all sides will happen via missile, bombers, and drones. In short, it would be chaotic and fast moving with destruction happening everywhere. The involvement of satellites is terrifying since they would be a prime target and would likely be one of the first major casualties.
There are over 3000 satellites in the Starlink network alone. How many anti-satellite missiles are there? Would taking out all those satellites be worth the cost?
After a few the rest would likely go down by debris. Not that it would matter as a world war would already have caused so much debris from destroyed satellites that nothing up there would work anyway.
Kessler Syndrome prevention pacts would function similarly to MAD policies, I imagine.