Can 5G “Wireless Fiber” Replace Fiber ?
Verizon CEO Lowell McAdam calls it “Wireless Fiber”. It is a reference to replacing the fiber connection to the home (FTTH) with a wireless connection. Verizon has stated it wants to implement the so called “last mile” with wireless instead of fiber in some regions.
If we move away from Verizon, and look at this as a more general question, what does it take for wireless to replace fiber? Will it succeed? Like most businesses, the devil is in the details.
What is the Job to be done
Anytime we look at a new technology with new business model, we must start with the problem that is being solved and what is the job to be done. There are two levels of jobs to be done here
- The end customer needs a broadband connection to the home that can support sufficient bandwidth for the ever increasing bandwidth hungry apps (e.g. Netflix, multiple devices, VR/AR in the near future)
- The service provider needs a technology that address this profitably and can be deployed very quickly in a neighborhood
Fiber has the throughput capacity to be able to support very high speeds (e.g. 1 Gbps), but fiber deployment to the home is painful with trenches to be dug at the customer premises. Think about laying a ring of fiber to every neighborhood upto the light pole, and then digging up the yard to drop fiber from the light pole to the residential unit. Doing this neighborhood by neighborhood for the entire country is not a task for the short term oriented or those without deep pockets. There is a reason even the mighty Alphabet is struggling: Google Fiber will lay off employees as it shifts to wireless. So fiber can meet #1 of jobs to be done, but not really great at #2
What does 5G Wireless Offer that Fiber doesn't
Wireless is less disruptive and more suited for connecting the “last mile” to the customer premise. 5G Wireless can solve the speed of deployment problem. Where it may take months to layout FTTH, wireless connections can be established in a matter of weeks. Wireless access points can be deployed on street furniture, such as light poles, effectively doing away the need for digging trenches to make a physical connection to the customer’s home. And a Customer Premises Equipment (a fancy word for the modem box inside a customer’s home) can be placed on the exterior or interior of the customer’s home.
Why hasn't this been done before with 3G or 4G? One of the reasons a 3G or 4G connection was never used to replace a fiber is because of capacity restrictions. 3G technology was never built for supporting very high speeds. And until now, 4G connection speeds never got anywhere close to 1Gbps. With Gigabit class LTE, this changes a bit, (see Telstra announcement), But Gigabit LTE is a step in the direction of 5G not a competitor to 5G.
This becomes interesting with 5G, which has been shown to support more than 10 Gbps in a lab setting. In order to achieve these high speeds, 5G relies on mmWave spectrum (25GHz to 100 GHz) that is needed to support the higher throughput requirements. So its possible that 5G wireless can provide speeds and the time to market where fiber has failed.
Unknowns of 5G Wireless Fiber that influence Economics
Sounds too good to be true ? Before we bring out the anointing oils for 5G, lets look at what are the unknowns. FTTH costs are estimated to be roughly $700 to pass a household and $650 to connect a household. For 5G to disrupt fiber, costs would have to a fraction of that of fiber. And 5G needs to be assessed just like any other business case.
There are several unknowns the industry is dealing with on the path to 5G.
Revenue Opportunity with 5G: For the use case of 1 Gbps broadband internet, Google Fiber has set the mark at $70/month. This is the expectation one should have even with 5G broadband. Note: I am not considering any incremental revenue made possible by low-latency applications possible on top of 5G. This will only further muddy the revenue picture and none of these are proven yet.
Spectrum cost of 5G: what is the cost of spectrum need to be in order for wireless fiber to be a viable business case? Can we pay the traditional spectrum costs and expect a positive business case with 5G?
mmWave Coverage: One characteristic of mmWaves is that they undergo larger path loss compared to lower band spectrum. Things like foliage, rain, buildings etc. impact the range of coverage. It is said, coverage could be in the range 50m to 500m per 5G access point depending on the path loss.
Cost of Rollout of 5G: What is the coverage we can expect for each radio? How many radios do we have to build out to blanket a given neighborhood?
Household Density: What is the number of household we can service per 5G radio. Is it more suitable for urban compared to suburban? How well would 5G work in rural geography.
Conclusion
So while 5G presents a compelling argument for fixed wireless use case, there are still factors to be ironed out, on this exciting new technology. If successful, 5G has the potential to bring trans-formative change to society.
