Needed: The Real Internet
Today's Internet is a far cry from its original promise. What will it take to restore it to all of the people that use it?
The Internet has moved a long way away from its original design goals of being an open and decentralized "network of networks." Today access is controlled by a few powerful interests exacting their tolls and control on everything their customers do while governments are dashing as quickly as they can to Balkanize the global Internet into fiefdoms of communication supporting party ideologies while cutting off access to communication deemed politically undesirable. The original goal of decentralized infrastructure has been reduced to concentrated conduits with on-ramps controlled by a very small number of corporations and governments. This vulnerability exists because of the insatiable desire of governments and private entities for surveillance and control of the people they should be serving and compromises in the design of the Internet itself that, ironically, were made to facilitate that surveillance and control.
We are blinded to this by service providers and powerful interests that place their own profit or power over value and utility to customers. Unless the FCC takes steps to stand up for itself (and get those steps past the courts) It will be a dismal end to the singular phenomena for humanity that the Internet promised to be. As with free speech, voting power, and so much else essential to free people's lives, capital and might have seized control and are subverting it to mercenary and political interests. This is not something new. The Internet has been moving in that direction since its potential started becoming apparent in the late 1990's and powerful interests began restructuring it to their ends. Unless you have the wherewithal and funds to peer fully, the Internet, originally "the network of networks" among peers has become mostly a product delivery service by and for a few large players who have been trying and looking for ways to control and meter what their customers can and cannot access while capturing and selling their customer's private communications for profit and dominance, with toll booths at every hop. It is this consolidation that has made the surveillance and censorship now rampant throughout the world so easy to accomplish.
It does not need to end there. The Internet will never be free while it has choke points that can be controlled by entrenched interests. There is no technical requirement that the Internet have service providers at all. Think about that. There are NO technical requirements that require commercial Internet Service Providers. They are the prime points of profit-making, control, and spying on people. We are blinded to this by powerful interests that want the Internet to be like cable TV so they can monitor, control, and parcel it out in pieces charging tolls to both providers and users for every service and click and denying access to those they don't want. It's about profits and control, not value and utility to users. It's interesting that the original military requirement for the design of the Internet was for it specifically to be decentralized and not have those choke points for obvious security reasons.
Leonard Kleinrock first proposed the idea of packet switching in his 1961 MIT doctoral dissertation. The concept was then developed by RAND Corporation researcher Paul Baran for the US Air Force in 1964. Donald Davies independently invented and named the concept of packet switching in 1965 at the United Kingdom's National Physical Laboratory (NPL). There was a lot of controversy in those early days over the idea of packet-switching networks. Packet switching breaks communications up into small blocks and attaches "to" and "from" addresses to each block before sending to any node within earshot. These nodes then pass the packets on to other nodes until they reach their destination where the communication is put back together - kind of like the post office. Electronic communication at the time relied on "circuit switched" networks where a "circuit" was established between the end points before any data was sent. Once that circuit was established communication moved back and forth as long as needed and then the circuit was dropped. This was the model of the phone system.
Packet switching was a radical idea that met with great resistance from established communication's concerns and it took small unknown companies working with academics to make it real. If I recall, AT&T actually refused to bid on the original military RFP's (Request For Proposal) to implement packet switching saying it was impractical and would never work. Of course, they had a huge investment in circuit switching at the time and this was a paradigm change that would make their technology seem archaic and inefficient. Later, after others showed that it did, in fact, work and money could be made of off it, they (and other powerful interests) became big players and the small enterprises that actually created the internet are today unknown to most people. (Does anyone remember BBN Technologies, originally Bolt Beranek and Newman Inc.?) Ever since those powerful interests have been trying to steer things back to their pay-per-minute (or pay-per-bit) and keep-everything-in-our-circuits-so-we-can-control-it model.
It's worth remembering that the bandwidth of the backbones of ARPANET and NSFNET were initially 0.056 megabits, about the same as the best download speed that dial-up modems eventually achieved. Some of you may remember the screeching noises those modems made when making a connection through phone lines to a computer that was actually wired to the Internet at the time. In 1991 some of that backbone had been upgraded to 45 megabits, similar to what a typical cable customer had a few years ago, and mostly remained that way with only 41 percent running at 2 gigabits by the mid 2000's. Until 1993 there was no World Wide Web. Back then there was no video nor any audio. Everything was text. Sharing a single still picture in a newsgroup or email involved breaking the file up into many pieces, posting them separately, and then downloading the many separate posts as file chunks and manually putting them back together again. A song might involve 30 or more pieces, video took hundreds and the download times could run into hours.
It's also worth remembering that the explosion of users from the mid 1990's through the early 2000's was mostly handled by small companies that often had started as bulletin board systems for communities of users. At first those bulletin board systems provided a gateway in their software to access parts of the Internet and later, as demand grew, those companies provided direct Internet connections via modems to their users over phone lines. These early ISP's often interconnected directly amongst themselves, bypassing backbones, especially if they had connections to the larger providers at peering locations. It was a different world and as it became apparent that the Internet was here to stay large corporate interests started planning how they could transform it to their liking.
When consumer broadband connections first appeared in the early 2000's via cable and DSL, these small providers couldn't compete against the cable companies and telcos except that the FCC ruled that the telcos had to make available for lease the copper wires running into homes and business. Most of those small providers jumped at the opportunity, signed lease arrangements for copper lines (and facilities at telco owned peering sites), and moved their new and existing customers to DSL broadband services. With the change in the political climate in the early 2000's, lo and behold, the telcos were able jack up the rates and otherwise make renewing those copper leases impossible for the small providers and in a matter of a couple of years, as the leases expired, almost all of those ISP's were wiped out and all of their customers easily picked up by the few large providers we have today.
The fact is the Internet is a network of networks and it is the consolidation of ownership, the dearth of interconnects, and the reliance on single point of control protocols such as DNS that make the Internet such a darling of the surveillance state and the large corporations. There is nothing in the design of the Internet that says it has to be that way. It is because of choices imposed by vested interests that it has become what it is. There are other ways to realize it. Even though it would be a large and diverse project, it is time to start making and implementing better choices. It is time we took back ownership. Much work has already been done by many people addressing the technical issues involved. We need to start putting all of the pieces together creating a new Internet that suits our desired future.
Although the IP protocol is still used, what passes as today's Internet is much closer in spirit to the closed services of CompuServe, AOL, Microsoft, and AT&T than the open and decentralized network of networks owned by users that was enjoyed in those earlier days and there is a huge and well funded effort by powerful interests to push that genie back into its bottle of consumerism, monopoly, and thought control. The kind of radical innovation that was the delight of users of those early years has been replaced by walled gardens of approved apps, devices, and services that are distinguishable from AOL or Compuserve only by their fancier graphics and use of more bandwidth. People have been relegated to being consumers once again. How many people do you know that make anything available on the Internet today that is not going through a corporate server? There was a time when user's controlled their own presence on the net and those corporate servers were the exception.
What is needed now is a back to basics look at creating a new Internet where the transport networks are wireless, cheap, simple, peer to peer, strongly encrypted, and built into the users devices eliminating the need for any carriers to provide transport services with their points of control and built-in surveillance. A free, open, decentralized, and secure global Internet that is wholly owned and controlled by its users is crying out for realization today.
As anyone who has worked in this area knows, there are many issues and limitations to address to make this so but I can remember the awe and wonder and challenge of participating as an IP node in the Internet as it was with a puny 1200 or even 300 Baud dial-up connection to a tiny 8 MHz PC back in the distant past of the early 1990's when naysayers were having a heyday and AOL and CompuServe boasted about being unperturbed by the nascent, tiny, and techie Internet of the day, calling it a "flash in the pan" that, like CB radio, people would soon tire of and come back to their "real" services. Much has happened since.
All of the routers and cell phones so ubiquitous today can already talk with with each other and route data through each other to more distant locations without the need for any service provider at all if only they had the software to do so. It's telling that they don't. In fact, re-purposing existing equipment was how many of the coops and home-grown providers originally built out their networks and provided service to people the large ISP's didn't see as profitable enough. I remember Verizon promising to bring broadband to certain unserved areas in a deal to obtain more spectrum for their cell phones and then later reneging with the CEO saying that the fine was only a few minutes of revenue whereas the deployment would cost much more. Your phone (or pc or whatever) could keep the packets meant for it and route packets onward to more distant locations through anyone within radio range who would then also do the same. Today this could be done with a "Routing Daemon" system module replacement and some small changes to the wifi and other protocols. The service would at first appear slow by today's standards but with time, technology, and political will it would quickly improve bandwidth and performance to today's expectations and beyond. Instead of using "pipes" owned by corporations and governments it would use the wide open physical space we all live in - an open 4 dimensional bandwidth space instead of metered linear toll pipes. Ultimately the space-bandwidth performance would be much better than is possible with today's techniques.
There are now 300 megabit radios with built-in routers that can be had for under $50. Around 15 years ago I experimented with Infrared and was able to connect nearby houses with 5 megabit links that I built then for under $10. Although the individual links may be slower than a fiber pipe, the aggregate bandwidth can be huge because there would be so many links covering the geography each carrying part of the load. Instead of a pipe-based point-to-point throughput there is a "sheet" based many-to-many transfer. The real problems are latency and routing. Because the links would be short there would often be many hops involved and each one involves processing and forwarding thereby adding time to getting a message across. Current routing protocols would also need to be modified to work in this new and dynamic environment. It would also be a good time to decentralize other aspects such as the DNS naming system. At first this would make real time two-way voice and video unreasonable, but in 1995 they were considered unreasonable on the Internet of that time. Look at it now.
Small "open" satellites, similar to the Starlink satellites, could address some of these constraints allowing for quick long-distance hops. What is needed is a legion of tiny "open" wireless routers in orbit all over the globe. I recently heard of an experiment putting a cell phone into orbit. A supply of these with the same software as their ground counterparts around the globe would do the job and be pretty inexpensive. The whole project could be completed at a cost less than what Facebook payed to buy Instagram.
One can also have very inexpensive "geosynchronous" satellites at far lower altitudes - by putting wireless routers up on tall poles or towers or buildings - no big rockets needed. They won't reach an entire hemisphere but can provide coverage out to 30 miles or so. Instead of satellites I call those "Terralites." There is so much innovation and opportunity possible here!
In addition, if each hop is individually encrypted by their owners and the routing spreads the packets around geographically there would be a huge increase in security - tapping such a network and putting the messages back together becomes extremely difficult if not impossible. The best part is that the hardware to do this exists and is already deployed throughout the world in the form of smart phones and wireless routers, it just needs the right software and the will to accomplish it. Another ten years of work could produce vast improvements on these simple beginnings.
The devices you use in the home should not have direct access to the Internet at large. All such devices should be behind a robust secure router, and not just the junk being sold today but something that can actually perform that role while providing secure VPN access to it's internal network to its owner. That is accomplished through software so there should not be additional hardware costs. Many of today's routers could be upgraded if the software were available. There is no reason for your thermostat, refrigerator, camera, or toaster to store your personal information on some corporate server or "cloud" other than for the owner of that server to gather, use, and sell your personal information. It is just another avenue to compromise your personal data. Devices that communicate directly with outside servers are open to hijack. This can be simply accomplished by making sure the device does not have a gateway address in its configuration - it won't be able to send anything outside of it's local internal network.
The other point of failure is the use of centralized services and facilities such as DNS and transport. The Internet, at its heart, is a distributed peer-to-peer system. Its design has been severely compromised by implementing critical services like DNS in single point of failure services and routing traffic through a limited number of physical routes owned by just a few entities. DNS should be redesigned as a distributed service with no discrete location or owner. The tools for doing this have been applied by peer-to-peer networks and block-chain applications with great success. The physical routing infrastructure needs to move beyond a few large companies to encompass all of the networks out there. There was a time when that was how the Internet worked but it changed around the turn of the century as governments and corporations saw opportunities to advance their own agendas.
This is just a simplistic start and accomplishing what needs to be done will take much time and work but mostly it will take a re-thinking of how we want the cybernetic universe we rely on more and more to serve us with the goals of personal security and integrity over profits and, in the words of former head of the NSA Gen. Keith Alexander, a "Collect it All" mentality.
DNS and other necessary services should also be distributed on users hardware eliminating another stranglehold. Social networks could exist distributed across user's equipment instead of on corporate servers, much like Bitcoin, shutting out another privacy theft. With point-to-point and end-to-end encryption built into its core and data traversing the countryside in pieces on random paths it would be extremely difficult to tap, monitor, censor, or control. No backbone or service providers needed. No monthly bills. The amazing things is that all of the hardware needed to implement this is already in place right now! You already own all the equipment you need to make this work. This new free open source global Internet could become a reality in days simply by people installing an open source app or replacing the firmware on their router. That's all it would take to start.
A new profitable industry could grow to provide ever better equipment for it. An open source software project could get it started.
It would finally implement the "universal service" that the FCC charges you on your phone bill but has never delivered (the money just goes to the telcos). It would be available to anyone for the one-time cost of the equipment or the installation of a free application on their existing equipment. You could say goodbye to your ISP and cable company for good. Of course there is no reason that it couldn't interconnect with the "metered product delivery" services that Verizon, AT&T, Comcast, and others want the Internet to be, and, in fact, would take a load off their systems - a problem which they constantly complain about and an excuse they use to limit service and charge extra fees. In fact, it addresses one their main arguments for eliminating net neutrality! They could charge tolls for their special services but you wouldn't have to use them. With "net neutrality" and security built into its core a new global Internet owned by everyone and no one would just appear out of nowhere while the "Old Internet" would be relegated to being a shopping mall on its periphery. It's time for truly innovative and creative people to step up and supersede what we have now just like the Internet once did to cable TV and the old phone companies.
I challenge the knowledgeable and capable people out there to put in the time and energy to craft and make freely available a new Internet, one that is open, free, secure, and real this time.