It’s a simple question how much spare fiber optic capacity should be placed in the network and where?
It can have a massive impact on your costs, too much and you’re wasting capex, too little and your opex costs go through the roof. Coming up with the right answer is not as easy as it might seem. The question is not how much fiber optic capacity do we need long term, but how do we get the capacity when we need it?
The Main Mistake: The wrong amount of spare fiber optic capacity in the wrong spot
Take a hypothetical simple 3 tier home run FTTP network (no splitters):
- At first glance, it looks like we could have 12x DPs (Demand Points) per MPT (Multiport), and 12x MPTs per FDH (Fiber Distribution Hub).
- However, the expected average take-up is 50% so that means up to 24x DPs per MPT.
- There is also an expected growth in DPs of 20%, so 20x DPs per MPT (24 / 120%) is possible. This places all the spare capacity at the MPT which will mean every new DP has only a drop to build so minimal construction and an overall minimal build.
- But the growth in DPs isn’t evenly distributed. Half of the growth is new clusters of 20 DPs, and the other half is scattered throughout the existing DPs. Now, the answer might be 22x DPs per MPT, and 11 MPTs per FDH.
- It is expected to take a number of years to achieve that take up. So if we take the time value of money into account, it will be better to delay some build until it is required. Now the answer might be 24 DPs per MPT, and 10 MPTs per FDH.
- However, the distribution of take up is highly variable. Some MPTs will fill up much quicker than others and so then the answer might be 20 DPs per MPT, and 10 MPTs per FDH.
As you can see, as more pieces of information get added to the calculation, the optimal distribution and quantity of spare will change. Add to this the other factors discussed above and it’s easy to see why some carriers don’t even attempt to run a model and thus make so many costly mistakes. Being able to model these variables and run simulations becomes a powerful tool in identifying how much spare capacity to place at each point in the network. It can also help understand what the cost implications are for different sparing rules, both for short and long term costs.
The other aspect to consider is the cost difference between the minimal capacity option and the higher capacity option. For example, when choosing between a 144f cable and a 288f cable:
- Cost of 144f cable now = $20/m
- Cost of a 288f cable now = $21/m
- Cost of a future second 144f cable = $10/m (Net Present Value)
If there is a 20% chance that a 144f cable will be insufficient long term, then the cost of installing a 144f cable first is $20 + $10*20% = $22/m, therefore the architecture is better to specify that a 288f cable is required.
How can we overcome these challenges?
As stated above some carriers don’t even attempt to run a model and thus they end up making a lot of costly mistakes.
Through automation and optimization across FTTx network design, we can re-evaluate how to simulate fiber optic capacity across a network. With detailed designs running at the core of a simulation, you can be far more confident across long term growth and ongoing strain across your fiber network.