Standardization of 5G is split into two phases. The first phase, which is scheduled for 2H2018, will focus on a subset of use cases and features. The second phase which is scheduled for December 2019, will meet ITU2020 requirements for 5G technology. Phase 1 will also focus on sub 6 GHz bands although carriers in Asia are pushing to include millimeter wave bands. The question is whether phase 1 will matter at all? Would operators bother with an intermediate step towards 5G, or opt to deploy the mature version when it becomes available?
In my view, operators will decide to skip ahead to Phase 2, bypassing Phase 1. The rationale for this is multi-faceted and has as much, or even more to do with financials, as with technology roadmap. Here are a few reasons:
1- No core. Phase 1 5G is a RAN upgrade over a LTE core. I have argued that the value of 5G is in the core and in how different access technologies can be combined. This is provided in Phase 2.
2- Limited improvement. LTE will continue to evolve providing operators with a roadmap to improve capacity in the sub 6 GHz bands. Phase 1 will offer limited differentiation over LTE in sub 6 GHz frequencies.
3- Limited spectrum. It takes on average 6 years to free up a new spectrum band. We are not yet at a stage to harmonize global bands for 5G as per WRC15 decisions related to mmWave. Spectrum availability will take us post Phase 1 availability.
4- Constrained finances. Operators invested a lot in LTE and will not invest in 5G unless they absolutely have to. Number of subscribers reached saturation. IoT connectivity has low incremental revenue and will rely on massive volumes to drive the bottom line. There has to be a compelling reason to deploy 5G. CFOs will press CTOs to justify these investments (should be fantastic arguments). LTE provides a flexible framework which is not perfect, but whatever problems it has, they are not mortal ones that will push carriers to deploy 5G quickly.
We have pushed against the limits of the physical layer. Major improvements in throughput will come from more spectrum, more antennas, and more cells. All are challenging and expensive propositions. Improvements in peak throughput will come primarily from getting close to the user (small cells) and carrier aggregation (more spectrum). Improvements in spectral efficiency in sub 6 GHz band is not going to yield major benefits – I reckon no more than 50% as a maximum over LTE, and perhaps the actual number is much lower. As a results, new deployments scenarios and processes are needed.
I don’t foresee ‘real’ 5G commercial deployments to commence before 2022. The 5G standard will undergo more trials than previous generations simply because it will have more variety of use cases and implementation options than previous generations. It will also have more interfaces which makes debugging and testing more challenging. Vendor lock and interoperability will become more acute than before.Until then, I expect we will be inundated with news of deployments that are in reality non-commercial market trials. Carriers will jockey to claim 5G services but with networks based on LTE and its enhanced roadmap. Marketing professionals at service providers will have a field day!