What is the best FTTH design?

As an FTTx network planning company we receive this question more and more. A recent LinkedIn discussion made it clear that this question keeps people busy - around the world. In this post, I want to address the question:

Does the optimal FTTH design exists?

Yes and No.

  • Yes, for each unique area of buildings/homes/businesses and streets/ducts/poles, and each set of restrictions (financial budget, optical budget, adoption, ...), optimized FTTH design rules exists. This optimization realizes the lowest cost FTTH design for that specific area.
  • No, optimized FTTH design rules for area A might not be the optimal FTTH design rules for area B.

So, why is there not a general optimal FTTH design? There are many reasons, we will describe only two. The biggest impact is related to the geography of the area, but also adoption has an impact to choose one design over another.

Impact of geography on FTTH design

Is the area urban, rural or something in between? Urban cities, for example, may choose splitters in the central office because the higher cost in feeding cable compared to decentralized splitters is small compared to the more efficient slot filling of the OLT cards. Low dense areas prefer to have splitters in the distribution point (decentralized splitters), as the savings in feeding cable have a bigger impact on the cost. These two extreme examples can be used as general FTTH design guidelines, however, not only density has an impact on this design choice.

How about the building profile? An area with only single dwelling units is totally different from an area with a few large multi dwelling units, although they can have the same density. It immediatly becomes clear that "density" does not describe which ftth design choice is the best.

Impact of adoption on FTTH design

Another impact is related to the adoption of the service. A short example:

  • P2MP (point-to-multi-point) design
  • 1:64 splitters in the distribution point
  • 64 homes connected to the same distribution point
  • expected adoption of 50%

In this scenario, (on average) only 32 homes per distribution point are activated, leaving 32 slots of the splitter unused - or 50% of the foreseen capacity at the POP (point of presence or central office) is unused. This is quite a waiste of investment. There are two easy solutions:

  1. use of 1:32 splitters instead of 1:64. Like this, you only need 50% of the OLT cards in the POP, but you need an extra feeding fiber per distributition point.
  2. increase the size of distribution points. Doing so, you will ensure that the splitter is more occupied.

What is the best solution? This will depend on (you can guess it) ... the geography of your area. If the distribution clusters grow, the amount of distribution cable increases, while the amount of feeding cable decreases. Depending on density, the impact can vary.

How should you design FTTH?

From the relation between density, building profile, adoption it must be clear that the FTTH design does not exist. People still use rules of thumb, combine them with years of experience to estimate the impact of all design rules. It is quite strange - to say the least - why someone would use such a strategy to select an optimal ftth design.

Better to know more first than less and pay a huge price after.

How should you solve this problem of optimal FTTH design? Introduce GIS (georeferenced data) and simulate the FTTH network. Simulating an FTTH network, results in GIS-files that show the location of trenches, feeder cable, distribution cable, ... as well as a detailed bill of material. Changing a design rule, enables you to compare the bill of materials and evaluate the impact on cost.

Introducing GIS helps you to really calculate -exactly- how many cables, trenches, ... you need for a certain FTTH design. Switching a design rule, e.g. another split ratio, will calculate a new design and you can easily compare the impact in terms of cost. This result was generated with the FiberPlanIT Simulator.


It must be clear that choosing the best FTTH design is difficult, or almost impossible as it depends on different parameters. However, introducing GIS data in the planning phase adds opportunities to simulate different rollouts. Add splitters in the basement and see if your total costs decrease. Change splitter settings from central office to distribution point. Change the size of your distribution points... Each simulated rollout has a bill of material as an output. Compare the bill of materials, and choose the design rule that fits your area.

Like what you have read?

If you read this article to the bottom, you must be interested in similar articles. Subscribe to our mailinglist to get them in your inbox! No spam, ofcourse - max 5 times per year.