Down a Privacy Warning Rabbit Hole
I've written about peer-to-peer technology before. It is essentially the underlying design principle behind DARPA's creation of the internet. Internet technologies were researched during the Cold War as a way of literally bomb-proofing communications networks, as telephone networks were essentially a client-server model. If an attacker wanted to cut off communications, especially for government, all they had to do was blow up the nearest telephone exchange to the facility they wanted disrupted. The internet inherently does what its name says, it inter-networks around centralised exchanges. When built to do so, at least.
The trouble is power (and power takes many forms, not just government or authority) likes its perview to have a baked-in central distribution point, so the telcos and tech bros who built the modern, public internet opted for the old client server model and built an infrastructure with a bomb-proof peer-to-peer backend. Then they put the clients on "spokes," making it impossible to connect via a hub then hand-off to a direct peer-to-peer connection.
While sections of this modern network have peer-to-peer connections, which is why the connections are so reliable and stable, there is literally no infrastructure that allows us to call our next door neighbour, device-to-device, via a direct peer network, at least as simply as making a client-server-client call. Or is there?
Mobile network devices, like our ubiquitous smart phones have the broadcast transmission power of a basic, low power walkie-talkie, between 0.2 to 1 watt, depending on the protocol and generation. Newer has generally become less powerful and more line-of-sight because of the desire to minimise our exposures to RF signals, which can be implicated (among many other factors of modern living) in some cancers. However, "micropower" transceivers can cover a radius of up to a kilometer or more at higher frequencies, and digital systems are very robust against noise and distortion, increasing the signal quality at the receiving end. We could remove the servers, the telephone exchanges/cell towers, as it were, and still be connected.
Unlike a walkie-talkie with a 1km range, a modern portable phone is a supercomputer in our pockets. It has the digital radio power to transmit a kilometre with very little electromagnetic radiation, but it also has the computing power to provide a network infrastructure for its surrounding area. It has an immasculated version of this already - mobile hotspot, the ability to provide a wifi network connection to friends travelling with you. It's a LAN party in your pocket, but the only reason it isn't a network hub and worldwide connection server is the telecommunications industry, data brokers and fearful government authorities. Don't worry, I'm still being rational, I'm not espousing a conspiracy theory, it's just normal, capitalistic self-protection. Not unreasonable for a small business, but bordering on creepy for big mega-corps.
So, what would it take to create a worldwide network of phones without cell towers? Sorry, but you're not going to do this with your current phone, the client-server model is baked in. It's cellular radio is locked down at hardware level to only be a client. What we need is an open source hardware device and software tool that is both cell client and cell base station. And that kind of tech has some heavy legal restrictions on public use. And not without fair reason, we don't want to clog all the radio bands with unregulated chatter, that would be a disaster.
What I want to see is a legitimate packet-hopping network, where I can make an encrypted phonecall home in much the same way a Facetime call works, but the link hops from nearby "client/server-in-a-pocket" devices, that may also be making similar calls home, to friends or to family. An iPhone or an Android hardware device has the "power" to be a 4 line cell connection, while also being a client at the same time. The personal cell network from incoming to next hop would be encrypted end to end. The data connections would be a bit slower than a "pro" cell network. Texting would seem just as quick, but video facetime might get a bit glitchy from time to time battery life would shortn a bit. All feels normal, right?
But here's the killer app, no centralised communications networks. Instead of swapping phone numbers, we'd swap a QR code or suchlike. Not really day-to-day human readable, difficult to reshare, because the QR would be timestamped, too. Yeah, I'm riffing here, but the power is in your pocket to do this, it just doesn't have the right comms chips to do it - it has client-server chipsets and code.
I hear you say, "if sombody's phone goes flat, doesn't that break the connection?" No, because it would use all the same sort of hand-off techniques of a client-server cell network. Another phone in another pocket could be handing off the call or two passing through it just before it runs out of juice. 1km range, roughly 1700 yards. Think of how many phones there are in a 1km radius these days. Yes, this requires as many people to adopt as have adopted the current model, so maybe there needs to be an added "pocket base station" that people could buy or build themselves. Maybe one of these could left at home, too.
Security? Wouldn't this make every phone hackable? No, the network chips can only hand off data until that data finds its recipient. End to end encrypted.
Wouldn't it take forever to find a phone across the world? No, this is literally what the domain name system does right now, every single second, except that this isn't a handful of DNS servers, like the current DNS, this is Distributed Doman Name Serving DDNS - it's a thing already, no single server keeps all the address, but they all keep all of the owner's addresses and recent paths for the few it they have in addressbook.
It has another layer of security, security by obscurity. Only those you allow can contact you. Block them, they can't find your connection. This is the best way to avoid being hacked, not having a findable system. In-person handover or secure emailed package. And before you say, "A phone doesn't have the processing power," they do. Apple are rumoured to be using their A18 iPhone chip in an impending MacBook launch. Another Apple, an LCII from 30 years ago, that was running in a faculty at University of Tasmania back then, was serving a CUSeeMe server for video conferencing. A Mac with Motorola 68030 chip from the 1990s. if an LCII could handle a facetime, an iPhone can handle a little bit of network routing in the background. An original RaspberryPi microcontroller running Debian could do it!
Client-server web browser will still give us the capitalistic surveillance state, but even a web server could run on an iPhone, so you could host your local community network and your friends would find it easy to join and would probably like your other friends. But maybe not your creepy cousin, but they can block them, and the cousin can't simply create a new account. It can still be compatible with the legacy network, too, but the little personal social network would be firewalled from legacy data harvesting.
This tech is possible, have no doubt. There are protocols and chips already that could build it, using some of those protocols might, um, "bend" laws a bit, but it might even be possible, considering how ubiquitous current phones are, to do it with a variation on WiFi hotspotting, although you might have to choose between the hotspot for your laptop on the train and being the infrastructure for the train you're on.
Oh, and one last thing. None of this exists yet. It needs new hardware or new apps, possibly both, either together or in parallel. There are platforms trying to create a world like this. have a look at the unPhone, a development kit peer-to-peer device out of the UK. More groundbreaking than RaspberryPi, but not as flashy and certainly not as old media as RasberryPi's ecosystem is. It's also not for the technophobe, but it's a tool that could build a lower-surveillance future.
Comments
Post a Comment