Monday, July 17, 2017

Trunking Fiber Decisions Now More Complicated

For the most part, fixed network internet service providers rightly have focused on access bandwidth (what is delivered to the end user or customer), and based distribution network decisions on what is necessary to deliver bandwidth at the network edge. For most legacy telcos, that has meant more fiber to home or fiber to curb (fiber deep, the only question being “how deep?”).

Cable operators, assuming use of copper media as the end-user connection, mostly have focused on ways to drive more fiber into the network, but without going “all fiber.”

As the need for small cells has become clear, all ISPs are asking different questions. Even if all trunking network decisions still are based on assumptions about end user bandwidth, the decisions are more complicated.

Even if all assume consumption is going to keep growing, at faster rates than in the past, so that both access bandwidth and trunking resources must increase, there now are more ways to supply that demand.

In some markets, the mobile network remains the “only” way to supply most of the demand. In other markets both fixed and mobile networks are potential suppliers. In a few markets there also are alternative facilities (cable TV networks), in addition to mobile and fixed. Also, 5G will add another option--fixed access--from the mobile platform.

In recent years, the ability to offload traffic from mobile to Wi-Fi access has been crucial. Going forward, the range of choices will grow. And that means more decisions.

Cable operators, for example, long have believed their distribution networks, and even consumer bandwidth, would become more valuable in the small cell era.

In part, that is because they can use their existing networks to support Wi-Fi access to mobile services.

Also, the thinking has been, the hybrid fiber coax networks have significant bandwidth available at the edge that could prove useful for supporting new small cells beyond consumer users at home. How the market develops, and how soon, will determine the extent of that value.

In principle, cable operators could provide wholesale trunking services for small cells, to third parties. Just how effective that strategy could become will be determined by where small cells are needed, the bandwidth those small cells must support, and how many such sites are needed.

The best scenario for a cable operator is “lots of small cells, but relatively light bandwidth demand,” as that supports maximum reuse of already-deployed capacity. The tougher scenarios are “high demand, few locations,” as that business case means potential customers can afford to  install new trunking fiber directly.

A similar sort of thinking now underlies the strategy Verizon uses to deploy its distribution fiber. The “One Fiber” architecture assumes a single trunking network that supports small cells, enterprise customers and consumers with gigabit bandwidth services, both fixed and mobile.

The difference is that the immediate driver is “fiber deep” trunking to support small cells (potentially many), that also has enough spare fibers to then support business customers. The small cell locations might then also be leveraged to support gigabit internet access for consumer customers, using radio drops.

Much depends on “how” that distribution fiber has been, or is, deployed. Extra dark fibers in cables will matter. The ability to use different colors of light will matter. The cost of overlaying new fibers will matter. The nature of demand (how many simultaneous users, for which apps) also affects thinking about the number of small cells, where they are located, and how much backhaul therefore is needed.

Up to a point, the more distributed, and the more numerous the small cells must be, the more optical fiber could be required. That is especially true for high-demand urban locations.

Beyond a certain point, highly-distributed demand means less need for distribution fiber.  In a very-highly-distributed consumption scenario (individual users in rural and many suburban locations, less fiber might be needed, since the density of demand is not so high, for any single user.

The example is gigabit 5G, as end user smart phones are among the best examples of highly-distributed consumption, but will be supported on a per-device basis at gigabit levels. In many use cases, the standard mobile network will suffice.

Similar observations can be made about consumer consumption at many home locations. Consumers who switch to Wi-Fi at home might mean ISPs are not required to deploy too much additional distribution fiber, beyond that needed to support gigabit fixed access.

The point is that decisions about trunking fiber arguably are more strategic than decisions about access fiber, since multiple “access” drops are feasible (coaxial cable, fixed wireless, direct use of the macrocell network.

No comments:

Many Winners and Losers from Generative AI

Perhaps there is no contradiction between low historical total factor annual productivity gains and high expected generative artificial inte...