By Tomas Jendel

Designing a fiber optic network

Designing a fiber optic network

Be it a car, or a camcorder, its performance, and operative qualities are results of the design. And, this also applies to fiber optic networks. Decisions – and oversights due to ignorance – made at the drawing table may affect operation and profitability for decades to come. A well-designed fiber optic network is key to operative performance, expandability, and, ultimately, the network owner's business. Here you can learn some fiber optic network design basics!

Network design

Designing a Fiber optic network is a truly challenging task where the designer needs to consider many different input conditions and perform many levels of optimization along the way. The network design phase is preceded by a network planning phase in which the overall business case for FTTH deployment is determined and strategic decisions such as architecture and basic passive material concept are made.

A good network design minimizes installation costs and provides long-term operation at low total cost of ownership.

So, what is a network design? Basically it is a set of deliverables such as area drawings, where the physical infrastructure such as cabinets, ducts and fiber cables is shown and schematic drawings over the duct and fiber network that will allow the installers to actually perform the installation and make the fiber network a reality.

The network design has a huge impact on both the Capex and Opex of a FTTH network. A good network design will thus not only minimize the obvious installation and material costs but also provide efficient upgrade and expansion possibilities and provide easy maintenance and a long network life.

Now, let’s explore what kind of input a designer will need to deliver an optimized design and what kind of decisions he or she is faced with!

Input, input and more input

It cannot be stressed enough how important it is to have access to all relevant input before the design work starts. Due to the inherent complexity of network design it is easy to make mistakes if the input data is flawed. Here’s an overview of what the designer needs:

Firstly, the planning phase must provide an area of deployment. This area must have a boundary that clearly shows the properties to be served. Also, it must be clear if there is a Single-Family Unit (SFU) on the property or if it is a property with many homes like a Multi-Dwelling Unit (MDU) or something else, like a business. This will affect the dimensioning of the fiber network.

Secondly, an architecture and design rules must be defined. Examples are network topology, i.e. Point-to-point or Point-to-Multi-point, spare capacity, redundancy requirements (not common for FTTH networks), suitable and approved deployment methods and access to existing infrastructure such as ducts or aerial pole lines. The available Right-of-Way (RoW) options, like existing infrastructure and local rules how and where trenching, plowing etc. can be performed, are critical for a successful FTTH project since most of the Capex is indeed installation related civil works.

Thirdly, the choice of passive material concept must be made. This includes decisions like whether to build with traditional fiber cables or to use micro duct technology or a combination thereof. Today there are excellent solutions on the market that give a designer a toolbox to design flexible networks that can easily be expanded in the future simply by adding spare micro ducts capacity day one.

Now that we have all inputs in place, let’s start the real design work!

Different stages of design

The design work is typically performed in stages, all with different input/outputs and purposes. The process to find the optimal design is typically iterative since there are so many parameters to consider. Today there are software tools available that based on an area and basic design rules can provide an optimal design quickly. It is, however, important that the design provided by tools is reviewed, particularly in the detailed design phase when documentation for network construction is developed.

To simplify the discussion, we assume that we are building an underground FTTH network based on microduct systems and microduct cables/air-blown fibers. This is a very common approach in many countries in Europe and elsewhere, but the design principle is the same with a traditional cable-based approach.

High-level design

As the name suggests the design is first performed on a high level. The starting point is to identify the locations and size of the nodes (Point of Presence, PoP, or Central Office) housing the active equipment such as Ethernet switches or GPON equipment. It is also not unusual that the node locations are given by existing infrastructure or by the network operator directly.

Next the area is divided into suitable sub areas that each has a fiber concentration point. All concentration points are linked to the node or nodes that serve the area, this is referred to as a feeder and distribution network. This network consists of a duct system and micro cables in our case that needs to be dimensioned to serve all properties in the sub areas. To cater for future expansions and developments and possible future wireless infrastructures such as 5G, it is important to add spare ducts in the network.

All properties to be served in a sub area are connected directly to the concentration point, this is the drop network. The concentration point is typically a street cabinet, an underground closure or a pole mounted cabinet/closure. As mentioned above it is an iterative process to optimize the size and number of concentration points. A common number of properties served from one concentration point is between 40 and 90 but this varies greatly with area density and other parameters. In the case of an air-blown fiber-based drop network a microduct infrastructure is built where it is possible to blow fiber units with, typically, two fibers from each customer premise to the concentration point or vice versa.

Based on the sub areas and the locations of the concentration points it is now time to design all duct routes. All targeted properties within the sub area must somehow be connected with a duct directly to the concentration point. Selecting the best duct routes is critical as mentioned before since most of the costs in a project are associated with civil works such as trenching and reinstatement costs. Some typical examples:

  • If it is expensive to cross local streets it may be better to trench on both sides of the street - even if the total trenching distance is greater.
  • If there are surfaces without asphalt or stones, i.e. grass, gravel or similar, it may be more cost efficient to trench a longer distance in such surfaces.
  • Using existing infrastructure such as main ducts in the ground or aerial pole lines can reduce costs significantly.

When all nodes, concentration points and duct/cable routes are placed on the map it is possible to develop a Bill of Quantities (BoQ). The BoQ consists of a material list and a list of services. The material list (or Bill of Materials, BoM) contains all materials required to build the network (ducts, cabinets, fiber cables etc). The services list contains all installation work such as trenching, digging of pits, placing cabinets, installing duct assemblies, blowing fiber cable, splicing etc. With all the quantities and pricing, it is possible to calculate the cost to build the network. Usually the cost is divided into a homes passed cost and a homes connect cost. The homes passed is valid for the entire area, but the homes connect is the cost to connect one single customer (on average).

This is a natural checkpoint to see if the business case developed in the planning phase will hold! If the cost is too high, it may be necessary to revisit the choice of area or find ways to reduce the cost. If there is a green light it is time to start the detailed design work.

Detailed design

Based on the high-level design it is now time to develop all documentation required to build the network. There are two main differences compared to the high-level design. Firstly, it is necessary to perform a field survey to make sure that the proposed design is possible to implement and secondly the documentation must be both more detailed and more complete.

In the field survey the proposed design drawing (often the high-level design can be used) is compared with the actual deployment area. The purpose is both to make sure it is possible to use the routes suggested but also to find ways to optimize the design. Typical changes to the design can be to switch to the other side of the road to avoid existing infrastructure, utilize grass/gravel surfaces more or access properties in another way.

Based on the field survey the detailed design is developed. Now it is necessary to show exactly which duct size and color to be used and exactly which fibers in a cable that are spliced in a certain cabinet. Typical detailed design deliverables are:

  • Overview drawing with all duct routes and cabinets
  • Build drawings with a lower scale for detailed construction work (scale 1:1000 or similar)
  • Duct schematic drawing – a logical schematic showing how the duct network is built up from the node all the way to each cabinet
  • Fiber schematic drawing – a logical schematic showing how all the fibers in all cables are connected from the node to each cabinet. This schematic can also show how the fiber cables terminate in ODF’s inside the node
  • Drop and premises design, i.e. where to enter properties and with what kind of duct. This part is often agreed with the property owner.


We have discussed what it means to design a fiber network. The fiber network designer needs to have a lot of relevant input to be able to develop a cost-effective design. Based on the input the complex and iterative process of finding the optimal design starts. But the time is well spent as the cost to build and own the network is very much depending on a good design.

Learn more about FTTH technologies and business from our experts.

Tomas Jendel
Chief Technical Officer and a former responsible for Ericsson passive fiber R&D. Tomas has 20 years' experience in developing fiber optic technologies. His knowledge and skills in systems architecture make him a true expert in optimizing network design for local challenges and various national standards requirements.


Keep me updated