Increasing Bandwidth Demand and Hybrid Fiber-Coax Networks
Global IP traffic will increase nearly three-fold over the next 5 years. This means that the need for bandwidth is already growing beyond current capabilities.
Why? Because operators have only a limited amount of spectrum within coax or fiber, and some of that spectrum is consumed by legacy and inefficient services or technology.
“You go back, Jack, do it again, wheels turnin’ ’round and ‘round”
— Steely Dan, “Do It Again”
Cable providers have a few tools to respond to bandwidth growth. They can perform yet another node split. There is also “node+0” or “fiber deep.” But these upgrade modes are expensive and unsustainable in a world of nonlinear bandwidth growth.
Atoms, electrons and photons
More fiber almost always solves bandwidth problems. When in doubt, lay more fiber. Fiber is king, there is no issue here with that assertion. In fact, the entire wireless spectrum can be contained in a single fiber optic cable.
The issue with laying fiber is cost. Installing outside plant (OSP, the physical cabling and supporting infrastructure and any associated hardware) is the most expensive undertaking for plant upgrades.
Why? Physics. Moore’s Law may govern microprocessors where the number of transistors doubles every two years. Computers, laptops, routers, and phones have benefited from this advance for over fifty years. You probably could not even read this blog on your laptop or smartphone without the geometric technological increases in transistor density.
Just as Moore’s law has predicted the consistent doubling of processing power, the less well-known Keck’s Law has mapped the exponential increases in fiber optic capacity. The benefits of the network effect of connecting and interconnecting those transistors together outweigh (in a nonlinear fashion) the benefits of singular processing power. Fiber optics provide the network connectivity of the Internet. That technology provides the access and speeds necessary for Netflix, HBO Go, and Amazon Prime Video, among others.
At the end of the day, even with the spectacular, nonlinear technology advances in transistors and optics there still remains the physical act of trenching and burying fiber optic cable. This involves the physical movement of atoms, not those of electrons or photons. In contrast to microprocessors and fiber optics, construction costs generally increase with inflation, despite many advances and efficiencies in outside plant construction.
Handling complexity
Upgrades are not only expensive but they are also complex. “I don’t have space on my plant to add more.” “Do I upgrade my plant to higher bandwidth, or do I split another node?” “Can I eliminate analogue video?” “Do I stay on DOCSIS 3.0?” “When do I move to DOCSIS 3.1?” “What is Remote-PHY and can it help?”
“If you choose not to decide, you still have made a choice.”
— Rush, “Freewill”
These are all extremely important questions our customers face on a regular basis. In the face of flat revenues and rising bandwidth demand, these decision points take can take on a fatalistic bent. Putting off an upgrade will lead to reduced customer satisfaction and an inability to remain competitive. To make matters even more urgent, the Cisco Visual Networking Index asserts that busy hour traffic is growing faster than average Internet traffic. Video is the source of the accelerated busy hour traffic growth. So, interestingly, Internet video shares a characteristic with classic video by having a “prime time”.
“You Can’t Get What You Want (‘Til You Know What You Want.)”
– Joe Jackson
So what is a hybrid-fiber coax (HFC) operator to do? Well, a network assessment is a great place to start. An assessment will help you baseline the current demand, as well as a map where bandwidth demand is growing. It is a view to the future of where bandwidth will be required months and even years out.
https://blogs.cisco.com/sp/increasing-bandwidth-demand-and-hybrid-fiber-coax-networks
