I wondered if traffic between the UK and Europe ran through the Channel Tunnel - which would be much harder to surreptitiously cut without being noticed and yes 80% of traffic does:
> Approximately 80 percent of the Internet traffic which runs between the UK and mainland Europe today travels through the Channel Tunnel
Props to that article for finding the gloomiest picture of the English station of the Eurotunnel. Brexit has been hard but it's not a nuclear winter just yet.
There are two risks: evesdropping and an enemy cutting your fiber.
Evesdropping should be protected against by encrypting everything. Have a device sampling traffic over the link and publishing anything found in cleartext. Embarrass users into encrypting their stuff. Then, on top of that, do point to point encryption of the data stream (which is super cheap to do - in fact nearly all high speed links do data 'whitening', which is encryption with a well known key - just encrypt with a secret key instead).
Resistance against the enemy cutting the fiber should just be to have lots of connectivity. Hard to cut 50 links all at once.
If cables are cut you can bet the US will retaliate which is why it hasn’t happened. You can also bet there are plenty of undocumented cables.
However I do wonder if we should just focus of getting space laser communications working to avoid some of these ground infrastructure issues, it seems much easier to cut wires than blow up space craft.
You'll not be able to get anywhere near the throughputs that we have with today's fibre communication networks using satellite communications, unless there is some revolutionary breakthrough (and by revolutionary I mean completely changing of how we understand losses in free-space, so extremely unlikely).
This means we will not really use satellite comms instead of fibre. That said there is definitely applications that benefit from optical satellite comms.
But even getting 1% of the throughput is enough to prevent modern life grinding to a halt.
sure - people won't be able to watch netflix, but basic things like credit card processing, sending emails, getting emergency alerts out, ordering an uber, etc. would all work fine with a 99% reduction in bandwidth.
To cripple a countries infrastructure, you need to cut 100%. And satellite links mean you can no longer do that.
In the 90's at one point one of Sprints connections between the US and Northern Europe had a problem, and a lot of traffic was routed south towards backups. The congestion caused made connections to the US nearly impossible but also severely hampered a lot of intra-European traffic for many hours.
While it may not entirely grind to a halt, taking out part of the capacity can still cause a lot of problems.
Possibly there should be emergency scenarios considered. Where government turns off video things (youtube and netflix) in case of major network degradation.
you had better hope people stop using modern webmails then, have you seen how insanely huge those pages are? we'd have to sit for HOURS to download them at dialup speeds
I remember in the mid 90s a fellow intern scratching his head about why a short video clip he'd emailed to his school hadn't made it through - it eventually did, but took about 36 hours to send through the local OpenMail -> internet gateway.
Every other email out the door did as well.
> To cripple a countries infrastructure, you need to cut 100%.
You could do a lot of damage with a few %, targeted in the right places, such as the so-called operational nets (as opposed to their business/IT nets) used by organisations that run critical national infrastructure. E.g. those that monitor / control power distribution.
I think you seriously unterestimate the amount of data society uses to function. If e.g. Australia would suddenly go to only 1% of it's throughput it would cause economic chaos and likely cause hyperinflation amongst many things. Too many things rely on data essentially being free (the development of data traffic prices over the last 4 decades is absolutely mind blowing), in fact even if data throughput prices would stop dropping that would likely have significant economic impact as business plans rely on those projections.
People have been saying this literally for 20 years now, yet even the fastest 6E wifi with compatible devices and router can barely break 1Gbps barrier in real life use(for me), and majority of the time I see it's much less than that. I mean it's great seeing real life 300-500mbps over wifi, don't get me wrong, but all this incredibly fancy(and very expensive) tech doesn't seem to beat a basic £5 cable quite yet.
You're right - its just too unpredictable. The bit that gets me is the hardware in laptops / devices. Apple seem to take some care in antenna design etc, and appear to be fairly fast / reliable, but many device manufacturers stick in whatever, if you get a link then great, call it a day.
You can only retaliate if you can attribute the attack first and that might not be possible 100% of the time.
The reason why this happened yet is because it isn't very benefitial for the effort it would take. Any rouge state actor can use the Internet in their enemy territory for far more impacting operations (e.g. propaganda, psy ops).
Terrorism would be a more realistic thing, but terrorists don't typically have submarines.
+1 Retaliation wrt. to "cyber security incidents" is seldomly a plausible deterrent, exactly because attribution is so hard. Consider a "simple" case like the Nordstream attacks; despite serious efforts there still is no conclusive attribution.
Protection and mitigation are usually much more practical defenses than the threat of retaliation.
Sometimes things like that are a demo of capabilities.
Countries like to demo their war capabilities to other nations simply as a kinda-threat.
Such demos help stop war happening, so I'm not against them. Wars only happen when two countries believe they will win a war. If one of those countries believes they will lose, they will just admit defeat before any actual fighting happens. If countries are aware of eachothers capabilities, there is less chance that both countries believe they will win.
> If cables are cut you can bet the US will retaliate which is why it hasn’t happened.
What, like Nordstream? The risk is that the US cuts the cable when it benefits them. It is still a risk. Ditto for who is it who is going to be using the cables to spy on people.
HP were were working with the british military on laser line of sight communication. If someone breaks the laser beam, you'll know about it. Certainly more secure than cables, but environment is a weak spot, smog, fog, leaves, etc etc.
I might be wrong, but I think with space there’s the potential for a Gravity (2013) style situation though, where if one satellite is blown up the debris could cause a chain reaction and take out a ton of other satellites?
You can point-to-point encrypt a single fiber segment at the physical layer. This is what classified networks do and why they're able to use the same public backbones as the Internet, even though most of the traffic is probably also TLS packets to begin with. This way, even the addressing metadata is encrypted.
I'd categorize fiber cutting under attacks on business continuity. So do other forms of denial of service attacks, like interfering with power supply. Yes, the easiest way to protect commodity infrastructure is redundancy. And while an instant total blackout is unlikely, continuous strikes are a problem (compare to electricity networks in Ukraine). Redundancy is costly, especially for long cables, where attack surface compounds into the costs.
Eavesdropping is an attack on business safety. So is injecting misinformation or altering messages. For data, encryption (and authentication!) mostly solves these issues at pretty low costs and I think that where this matters it is mostly covered.
And who is going to do that ? The countries where cables are landing and who benefit the most from that eavesdropping, like the UK (in partnership with the US) and France ??
Important to note that the EU Commissioner for the Internal Market is an ex Telecom executive by the name of Thierry Breton who has demonstrated willingness in the past for aggressive regulation to help the telecom companies.
As a recent example, he has advocated levying fees on "Over The Top" services, ie Netflix etc that goes into a fund to pay for the Telecom companies investment into infrastructure.
On one hand, I don't think levying OTT fees make much as telcos ought to have been already paid for the bandwidth.
On the other, I also think OTT companies, and in particular VOD companies like Netflix really have no reason to rely on undersea cables in the first place. Simply put: neither netflix, nor any of their customers are located underwater and 99% of their offering is static content. Which means that distribution necessarily relies on CDNs and it's in everyone's interest (Netflix, the telcos and the customer) to access the copy located nearest to the viewer.
So in principle, content should never be delivered to customers using undersea cables and indeed this typically happens only when people use a VPN (to bypass somewhat silly IP licensing restrictions designed by the content owners).
Netflix needs underseas cables (well, transoceanic links) to serve the critical dynamic content / account state keeping and also to transfer content to CDN nodes. The catalog of content changes regularly.
Yes, but the Commission takes decision as a body of 27, not by individual Commissioners. What he might say and what the commission puts forward is different .
Not to mention that the commission in all important matters only makes proposals which are then depending on the topic adopted either by the council (ministers or heads of state of all 27 EU countries) or by council and Parliament.
OTT refers to the service being delivered via L3 IP, hence the ISP/Telcom can't block it or charge for it (as they would prefer to in the intests of making more money). So the content is flying "over the top" of their gate.
(Telcos didn't want the internet to exist, with its service neutral model : they wanted to sell "value added" services over their wires).
Perhaps EU should reclaim the Mediterranean sea and the Baltic sea and turn it into land (perhaps also part of the North sea between UK and EU). That would prevent Russian subs from messing with any cables AND it would protect against flooding due to climate change.
You could over engineer it to be more than just a dam; build any hydro-electric power off of side-canals, and the main dam could be two dams that develop into a full land bridge of considerable girth.
> Another proposal for which the EU would have available funding is the Far North Fiber, an internet cable to connect Scandinavia to Japan via the Arctic to avoid major choke points like the Suez Chanel and the South China Sea, revealed by EURACTIV last October.
> Approximately 80 percent of the Internet traffic which runs between the UK and mainland Europe today travels through the Channel Tunnel
https://www.datacenterdynamics.com/en/news/colt-completes-th...