NG-PON2 and 5G wireless seem designed to work in tandem, with NG-PON2 serving an important role in 5G backhaul/fronthaul. Yet both make similar customer promises -- ubiquitous broadband providing enhanced video with next-generation services for virtual reality, Internet of Things, and so forth. So NG-PON2 and 5G also compete.
This makes for interesting analysis and that's what we did in Communications Industry Researchers (CIR) "Current and Future Markets for PON, the Evolution to PON 3.0 Technology 2017-2026." We conclude, when costs are right, NG-PON2 will become dominant in 5G infrastructures. If NG-PON2 can prove itself in 5G trials, the market for NG-PON2 for 5G backhaul/fronthaul will reach over $890 million in 2022 going on to almost $1.1 billion in 2026. But the jury is out on NG-PON2 versus 5G as a platform for advanced services.
NG-PON2 has the best technical fit for 5G backhaul: 5G's bandwidth and latency requirements mean that more of the mobile backhaul/fronthaul infrastructure will use fiber than in the 4G LTE era. But for now, only Verizon has bought into NG-PON2 in any significant way.
Passive optical networks (PONs) offer the lowest costs for fiber in 5G infrastructure. PONs combine these attractive economics with latencies -- 1 to 7 ms -- adequate for most of the services 5G will offer. Just adequate: futuristic 5G services -- virtual reality, for example -- have more scope with 1 ms. For now, 1 ms latency can be achieved only over Carrier Ethernet or wavelength division multiplexing (WDM) at high costs.
NG-PON2 and XGS-PON are the only PON varieties proven both in latency terms and in terms of data rate. NG-PON2 offers four full-duplex 10Gbit/s feeds. There is an NG-PON2 version that provides 80 Gbit/s capacity in each direction. By 2020 an NG-PON3 with 25 Gbit/s channels may have appeared. NG-PON trumps all other PONs because it also offers:
Capacity management based on its tunable optics
Built-in equipment protection using wavelength switching
Wavelength channel bonding
Coexistence with RF video.
All this shows how far we have come from the old FTTH PONs. NG-PON2 and successors will be more than a low-cost way of deploying fiber for mobile infrastructure. It is a complete solution providing the management needed for a service provider platform.
NG-PON2 costs must come down: Despite being a PON, NG-PON2 is still expensive (by PON standards). This reflects the high costs of embedding tunable optics. We anticipate there will be improvements here, using lower-cost thermally tuned DFB lasers or possibly DBR lasers. NG-PON2 costs also will drop by reusing existing PON fiber, although there then is a need to add coexistence elements and splitters.
Verizon dominates NG-PON2 deployment at present, and Verizon's experience will shape the market for NG-PON2 for 5G backhaul/fronthaul. Verizon could be the NG-PON2 customer that pulls NG-PON2 out of that old chicken and egg problem: no volumes without low costs, no low costs without volumes. This will be good news for vendors like Cisco, Nokia-Alcatel-Lucent, Huawei, Calix, ADTRAN, Ericsson and others that see NG-PON2 as a long-term opportunity.
Verizon may yet discover NG-PON2 is not a success. 5G infrastructure will then trend toward XGS-PON, where low costs seem a better bet. Unwilling currently to take the technological risks of NG-PON2, AT&T and Vodafone plan to roll out XGS-PON, although they too plan to shift to NG-PON2 if it becomes more competitive.
NG-PON2 vs. 5G: While NG-PON2 may or may not be the solution to 5G backhaul/fronthaul infrastructure, there seems to be a looming clash between 5G and NG-PON as future platform for residential broadband services:
On one hand PONs are mature for broadband delivery. NG-PON2 might therefore be considered the natural progression for residential broadband. Costs are clearly a challenge with NG-PON, but the data rates and latencies NG-PON2 offers are adequate to provide the improved video quality and download times expected by carriers and customers.
Yet the main purpose of 5G is provide a wireless platform for similar broadband services to those that NG-PON2 will also provide. These include not only enhanced video but services supporting IoT, virtual reality and augmented reality services.
So, while NG-PON2 and 5G are synergetic
from the perspective of infrastructure, they are competitive
from the perspective of service offerings. It is tempting to predict NG-PON the winner since it is a mature technology -- it could be two years before 5G standards are completed. Nonetheless, recall how many services have switched to wireless in the past two decades. Therefore, while the jury is out, a plausible future is that NG-PON2 and its successors provide much of the infrastructure for 5G, but PONs begin to retreat as the platform for FTTH.
— Lawrence Gasman is the founder and president of CIR and has been tracking commercial opportunities in high-speed networking business since 1985. Over the years, he has covered both the optical networking equipment and optical components market and his recent market analysis work has focused on optical data centers, embedded optics, 5G infrastructure and quantum networking. Follow him on Twitter @cir-inc