When I first started working in fiber optics 40-plus years ago, the world was quite different. Computers were big rooms full of cabinets used by businesses and monitors had cathode-ray tubes, just like the televisions back then. Phones were connected by copper wires, except for some long-distance networks over satellites. Computers mostly did not connect to each other, except by people carrying around disks or tapes—unless you had access to secret experimental government networks.
If you asked me then to predict what would be happening in 2020, I would have had no ideas about the internet, a world connected with fiber optics, wireless smartphones or computers on every desktop and in every pocket. Large, flat-screen monitors and TVs were strictly science fiction then. All the world’s knowledge available on a hand-held device—no way, except in “Star Trek!”
How does one “future-proof” anything when technology evolves so quickly? Years ago, I had a training course on forecasting. The instructor made the point that, “All forecasts are wrong from the moment they are made.” But he taught us to carefully analyze the past, because that’s all we had for real data, and try to see what that data implied might happen in the future. Limit yourself to trying to predict the near future, because the farther out you go, the less accurate the forecast will be. And, of course, keep regularly adjusting your forecasts.
So let’s get started analyzing where we are now and see if we can predict the near future, starting with the telecommunications outside plant, the heart of our communications backbone. All backbone communications today, except in some very rural areas, are on fiber optics. Even those areas where wireless or satellite communications are used for cost reasons are eventually tied into the fiber backbones.
If you are designing or building a fiber optic cable plant, you will probably be planning on a 20-year lifespan for the facility, although it’s likely to last much longer. Over those two decades, the communications requirements along that route should be expected to grow exponentially, so how do you plan for that?
First, when doing underground construction, add in plenty of extra fiber ducts—the “dig once” policy. Using microtrenching to install multiple microducts and microcables, one can start with 288 fiber cables that are hardly larger than a pencil and add more when needed without additional construction. We see this everywhere now. Because fiber is “cheaper than kite string or monofilament fishing line,” as one friend in the business often remarks, it seems prudent to install a lot of fiber now. That long lifetime means future network expansion cost will mainly be in communications equipment.
What fiber should you install? G.652 has been the favorite single-mode fiber for, well, forever. Its specs keep getting better as it is able to handle more speed, distance and wavelength-division multiplexing, but it’s not really different from the single-mode fiber introduced in the mid-1980s. It should meet your needs for several more decades.
What about aerial cable plants? Personally, I worry about the future of aerial cables. I’ve seen too many instances of shoddy workmanship or aerial cables destroyed by weather. Sure, it’s cheap now, but does it have a future? Go underground if possible.
Premises cabling is another story. Cabling for local area networks (LANs), distributed antenna systems (indoor wireless), building management systems, etc., faces a mishmash of standards and incompatible networks.
Some people might advise installing a Cat 6 copper cabling system today that can also be used for powering LED lights, ceiling fans and computers—but not me. Power over ethernet became a problem when the applications grew beyond powering Wi-Fi access points and voice-over-internet protocol phones based on liberal interpretations of the technology.
Connectivity inside buildings should be based on wireless, mainly Wi-Fi, for most users who mainly depend on mobile devices and occasionally need wired ports for high-performance computers. The classic structured cabling network is rapidly being replaced by the simpler, less-expensive passive optical LAN (POL), based on fiber-to-the-home technology using single-mode fiber. (See my article on ecmag.com this month for more on POLs.)
Is multimode fiber obsolete? Maybe “obsolete” is too final a term, but the world is now focused on single-mode for every application. Multimode is struggling to keep up with network speeds, much like Cat 5/6 did in the past. Electronic components used to be much more expensive for single-mode networks, but are now considerably cheaper, making single-mode networks a much better choice and, in fact, the only choice if future-proofing is one of your concerns.