Over the past two weeks. I've written about the privacy implications of the Internet of Things (IoT), with respect to both data security and surreptitious collection of information. Today we consider a more mundane, but no less practical, aspect of the IoT explosion: bandwidth.
Consider this: Cisco IBSG predicts there will be 25 billion devices connected to the Internet by 2015 and 50 billion by 2020. Finding each of them their own connection to the internet won't be a problem, thanks to the switch we made a few years ago from IPv4 to IPv6, which gives us more IP addresses than there are atoms in the Earth. No, the problem is processing all that traffic on the network simultaneously. Suppose the number of automobiles on the roads were to double over the next five years. What would happen to our already-crumbling infrastructure? Even if we all had private garages for our cars, we'd be left sitting in them, waiting to get out of the driveway.
This is not a new concern. As early as 2011 (and probably even earlier), IoT forecasters observed that "there may not be enough space in the public airwaves for all the things that want to chatter with one another."
Google Developer Advocate Don Dodge repeated the prevailing wisdom last year when he concluded that "we're going to have to ...build a brand new network" to handle all this traffic. Various solutions have been proposed, from "a special low-bandwidth cellular network" to "Wi-Fi-like hotspots that are themselves linked via fiber." The French company SigFox is already testing a new type of cellular network just for "things" in the San Francisco Bay area, using the unlicensed 915-megahertz spectrum band most often used by cordless phones.
If successfully implemented, such low-range interactions will add local flavor and variability to our digital interactions, allowing individual businesses to better customize the exchange of information with customers. It will also make it enormously more difficult for law enforcement, advertisers, or anyone else to efficiently search for particular users or transactions.
It seems quite likely that this practical limitation on existing networks is one of the driving forces behind the effort to provide truly global internet access over the entire planet--what I describe in my recent book as the "panternet." Google and SpaceX are building a satellite network for this purpose, while Facebook wants to blanket the world in broadband using high-latitude balloons. Although the stated (and laudable) goal of these projects is to extend internet access to the billions who don't already have it, surely the networks will also enable a new class of IoT devices to connect as well.
Indeed, these bespoke networks (whether locally generated or globally broadcast) may have no need to connect to the "internet" as we know it, beholden as it is to undersea cables, ICANN oversight, and the looming tides of federal regulation. To the contrary, these may be new "Wild Wests" of private autonomy, much like the internet itself used to (inaccurately) characterized. At that point, IoT devices may become a passport to a decentralized digital world the likes of which we have not yet experienced.
