Building a Fiber Network for the Next 20 Years

The pace of fiber network deployments is increasing as service providers race to accommodate the growing bandwidth needed for next-gen data transmission. The demand is widespread, coming from several fronts, including data centers, enterprises, and the metro backbone, as well as to fuel services like 5G.

With no slowdown in sight, it’s time to look ahead and consider what the fiber network of the next two decades will require. Deciding how much fiber and what fiber counts are needed in any given network is important from both a technology perspective and an operational one. While determining the level of capacity needed for a given route is sometimes made on the ground, high-quality networks are planned based on the types of customers that will be served and the bandwidth they may require. Building a scalable, high-capacity network is about planning for the end users’ present-day needs while also predicting customers’ escalating needs for the next 20 years.

So, what is the industry’s answer to this growing demand?

Fast-tracked upgrades and overbuilding

The tsunami of high-bandwidth applications is causing fiber service providers to fast track deployments. Planning for both today’s needs while anticipating future requirements requires upsizing fiber counts as well as building pathways that can easily accommodate overbuilds and upgrades.

And it’s not just the amount of fiber that needs an upgrade. Today’s technology is changing so fast that new generations of hardware are introduced every few months. As low-latency requirements for new services like 5G increase, next-gen routing and switching equipment also must keep pace with the latest service level requirements.

As the saying goes, past behavior – or, in this case, usage – is the best predictor of future usage. In today’s age of (hyper) connectivity, however, the status quo is a moving target. Instead, it’s best to plan networks for the eventuality that more throughput will be needed than originally anticipated. That means overbuilding the network for these services now or designing the network and preparing the pathway, so it's easier to deploy more fiber later.

Fiber counts matter

Fifteen years ago, the workhorse of a network was a 96-count fiber. Today’s network standards call for 288- to 864-count fiber and upwards. In the case of data centers that also are seeing an uptick in bandwidth requirements, for example, 3,456-count fiber might be used.

When looking at the overall cost of construction, the fiber itself is a small percentage of the total build, so overbuilding to ensure future capacity makes business sense. It’s much more expensive – and disruptive – to dig up a roadway or sidewalk to install more fiber later than it is to run more fiber during initial construction.

The key is striking the right balance. There is such a thing as overbuilding too much. If 48 fibers are needed to serve an area today, for example, putting in 1,728 fibers is probably overkill. Many fiber service providers are formulaic in their approach with new builds without committing to huge overbuilds.

Changes in fiber

Changes in fiber in the past few years have made it more efficient and sustainable for future needs. While the purity of the glass has been static, with G652 remaining the standard, the physical dimensions of the fiber have changed, thanks to new technologies and techniques that reduce its interior mass. Newer fiber has removed the strength member within the fiber, and “glue” is instead used to hold fiber strands together, reducing the size of bundles by about half. A tube that used to be an inch across, for example, is now a half-inch and contains the same or more fiber.

This is a huge advancement for the industry. It’s become much more mainstream to see 432- or 864-count fibers using this denser composition. This is especially true in areas where there is a critical Department of Transportation crossing or a leased third-party system where capacity is needed, but space is limited.

Planning for next-gen services

Engineering scalable networks to meet quality of service (QoS) standards takes more than higher fiber counts. Using clear metrics around capacity requirements for specific instances – such as lateral to a single tenant, ring, or backbone – fiber service providers are able to manage the frequency of upgrades and maintenance, reducing disruption to the end customer. The result is consistent and reliable uptime, month over month.  

For the fiber service provider, planning for the next 20 years means building in enough front-end capacity and creating routes that are strategically designed, constructed, and permitted in a way that supports their longevity and ability to be easily upgraded. This creates ongoing QoS that helps customers succeed and scale faster. Ultimately, if a service provider builds its network correctly, it will fade into the background. After all, a network that is always on means a job well done.

Joe Pellegrini is COO of Everstream.