Never has there been so much hype and attention paid to a new wireless standard than with 5G. 5G has generated a lot of interest, due to its potential transformational impact on both consumers and businesses across the globe. Has the hype been overdone? Let’s look at where we have come from, where we are today and speculate a bit as to what the future may hold.
How We Got Here
A little over 10 years ago, Apple introduced the iPhone and opened our eyes to the potential of smart devices combined with wireless broadband data. In 2016, Cisco published their Global Mobile Traffic Forecast, estimating that over 1.5 billion smart devices were sold globally. The report also estimated that by 2021, the world will consume over 49 exabytes of data per month—a 7x increase over the usage in 2016. The acceptance, adoption and pervasiveness of smart devices astounds and has been, in and of itself, transformational. 5G aims to go further. Broadband wireless data will continue to draw attention, and the world’s standardization bodies shaping 5G have taken notice.
At the outset of the 5G standardization kickoff, the 3GPP outlined three key performance metrics (see Figure 1). The 3GPP defined the enhanced mobile broadband (eMBB) use case and attached a performance target of greater than 10 Gbps peak data rate to expand broadband data services. In response to several industry groups, the 3GPP also set key performance metrics for ultra-reliable low latency communications (URLLC) and increased connectivity, setting the stage for billions of connected devices for massive machine-type communication (mMTC).
Today’s wireless standards do not and never have addressed latency and broad connectivity. Latency is quite important because lower latency will not only improve the common data experience for users but create new applications that rely on fast network response. Low latency and, specifically, deterministic low latency responsiveness provide the foundation for “control” applications over the network. Combining robots, drones, cars and other devices that “move” with low latency wireless communications makes controlling these devices from a remote location possible, potentially impacting construction, medicine, manufacturing, retail services and safety. Latency in this context also includes delivering timely information from the cloud or deployed sensors to the brains of these devices, so that decisions can be made on the fly to enhance safety. In this case, the data is delivered real-time, and the control mechanism is deployed on the device.
In 2015, when the 3GPP kicked off the 5G standardization effort, the group outlined the timelines and key performance objectives for this new standard. The 3GPP stated unprecedented guiding principles for the definition process to follow. First, 3GPP broke compatibility with prior releases, setting a goal of forward compatibility. By breaking with LTE and prior-generation standards, the 3GPP opened a path for innovation to meet these very difficult objectives. Second, the 3GPP divided 5G into phases. The first phase, or Phase 1, focused on mobile access below 40 GHz and set a framework for Phase 2 to investigate spectrum above 40 GHz. In all, the 3GPP has been working on 3GPP release 15, also known as 5G New Radio (NR) Phase 1, with an expected release date of June 2018 (see Figure 2).
Where We Are Today
As I write this article, the 3GPP is closing in on the first draft of the physical layer of 5G NR Phase 1, targeted for December 2017. This first draft is critically important, as it establishes the foundation upon which semiconductor, device, infrastructure, test and measurement and other wireless ecosystem players will plan and build their businesses. Until this point, the development has evolved using system prototypes for field trails with service operators. With a firmer standard in place, the players can develop tangible plans and targets for product and service rollouts.
Figure 3 Initial 5G deployments will use the existing LTE radio access network and EPC (a). The NSA specification includes architectural options using the new 5G core (b) and (c).
Interestingly, there have been announcements regarding the availability of 5G technologies—specifically by Intel and Qualcomm—and these early developments are intended to seed other companies to drive adoption. It may seem strange to announce products compliant with a standard before the standard is final, but the shape and structure of 5G NR has been crystalizing for several months. “Final” solutions will inevitably need some tweaking to meet the standard; however, progress toward the creation of the ecosystem has already started with a path toward commercialization.
Service operators have announced 5G plans in all shapes and sizes: SKT and KT are gearing up for 5G trial services to accompany the 2018 PyeongChang Winter Olympics in South Korea. In the U.S., Verizon has aggressively purchased spectrum in the 28, 37 and 39 GHz bands and driven the development of the 5GTF standard, primarily for fiber to the premises (FTTP) applications. Verizon has been trialing pre-commercial equipment in 11 cities in the U.S. since the beginning of this year and announced plans for initial commercial deployments in 2018. T-Mobile, the big winner in this year’s FCC auction, won 31 MHz of spectrum around 600 MHz and announced plans to build a “nationwide” 5G network using their newly purchased spectrum. Sprint has approximately 120 MHz of spectrum in the 2.5 GHz band and has been working with Qualcomm and SoftBank, its parent company, to plan 5G rollouts in 2019. Meanwhile, AT&T has announced plans for IoT services in spectrum it currently owns and acquired FiberTower to obtain licenses at 24 and 39 GHz.
Since the 3GPP kicked off the 5G standardization effort in 2015, the mMTC use case has been deprioritized. The 3GPP continues to evolve LTE; in release 14, the 3GPP made several enhancements to LTE specifically targeting the mMTC use case, with development of the NB-IoT and LTE CAT-M standards. The mMTC use case elevates connectivity as a goal, driving device manufacturers to incorporate wireless capabilities into many devices not previously connected, expanding their utility. We have seen a glimpse of the possibilities with new IoT devices, but there are significant challenges: there is no pervasive or ubiquitous wireless IoT standard. As such, there are challenges with interoperability and seamless connectivity to infrastructure and smart devices. With the 3GPP addressing the mMTC use case in release 14 and delivering a comprehensive and widely supported standard, time will tell whether further enhancements are needed in a future evolution of 5G.
What Will 5G Look Like?
As noted, the 3GPP plans to finalize 5G NR Phase 1, 3GPP Release 15 by the end of 2017, with the ASN.1 ratification in June of 2018. The 3GPP has started on the path of defining the transformational radio access network by including wider bandwidths, essential for faster data rates. New spectrum has been identified to deploy these wider bandwidth systems. The 3GPP has also reduced the symbol timing compared to LTE, to enable shorter transfer time intervals (see Table 1). In addition, the 3GPP has aligned on a self-contained subframe, which enables transmission and reception in a single subframe for time-division duplexing (TDD) systems. With this initial work, the 3GPP has addressed faster data rates and lower latency. Perhaps most importantly, a new, flexible numerology will enable operators to accommodate different types of devices and support diverse use cases.
For mmWave, the 3GPP has identified specific frequency bands and incorporated beam management and control for phased array antennas (PAA). Although the stage is set for mmWave deployments, many practical challenges remain for widespread adoption (more on this later).
The 3GPP defined two main network architectures for Phase 1. With the non-standalone (NSA) architecture, the 5G NR uses the existing LTE radio access network and evolved packet core or EPC (see Figure 3a); NSA includes two additional options (see Figures 3b and 3c). The second main architecture, named standalone (SA), uses the 5G NR and a new 5G core (see Figure 4). NSA enables operators to offer 5G services sooner, taking advantage of the existing infrastructure to deliver services in the short-term, since investments for SA are expected to be much larger and will take more time. In the standards meetings, NSA has been a focus because of the immediate opportunity and, perhaps, a narrower scope. SA will deliver more 5G benefits than NSA and will surely improve data rates and latency to be much closer to the 5G targets. The 3GPP is targeting December 2018 to wrap up Phase 1, release 15, which will include both NSA and SA.
The 3GPP has accomplished much in a very short time. In the short-term, 5G deployments below 6 GHz may look a lot like LTE on steroids, i.e., faster data and lower latency. The first NSA deployments may provide noticeable performance enhancements over LTE, and the lightning fast data speeds will likely appear when network operators deploy mmWave technologies and the new 5G core network, needed for SA operation. What is clear is that this is just the beginning. Future evolutions and iterations seem inevitable.
As the 3GPP finalizes the formative 5G specification, the path forward is not unimpeded. 5G has the potential to be a “game changer,” but the transformational impact must come with extensive help from a diverse set of players. Potential challenges exist in three high level areas: mmWave, network topology and ecosystem.
The 3GPP chose to incorporate mmWave technologies into the standard due to the scarcity of available spectrum below 6 GHz. More spectrum equates to faster data speeds. Although the 3GPP will specify 5G technologies for use in spectrum below 6 GHz, the 3GPP is relying on mmWave, with its copious spectrum, to meet its goals for the eMBB use case. Over the last couple of years, several researchers have prototyped mmWave systems extensively, but the early prototypes were big, bulky and used very new technologies such as PAAs.
PAAs overcome the free space path loss associated with mmWave transmission and reception using multiple antenna elements and beamforming to enhance gain. With their benefits, PAAs also pose system challenges, as the control of the beams must be incorporated into the standard and, more practically, into the software deployed on these systems. To support mobility, the protocol software must switch the beams in less than 200 ns to maintain the link, requiring fast switching technology in the antenna assemblies and the software architectures that program them.
The testing of PAAs and the systems that incorporate them is being investigated and poses new challenges for the test and measurement industry. As PAAs are often integrated with their transceivers to minimize loss, cable access to these modules and the arrays that incorporate them will not exist. Over-the-air (OTA) techniques for testing PAAs are being explored by several companies, with proposals submitted to the 3GPP RAN4 working group for incorporation into the standard. OTA introduces new variables to the test equation; most significant is the need to minimize test time and test cost. Test and measurement companies must deliver fast, cost-effective solutions to the wireless industry to facilitate the development of the mmWave ecosystem.
Even with PAAs in both the user equipment (UE) and infrastructure (i.e., gNodeB), mmWave propagation is limited, even at the lower mmWave frequencies. Denser deployment of the infrastructure is a foregone conclusion that will likely signal more costly rollouts of the technology and systems. To address the density challenge, researchers are exploring new techniques for mesh and integrated access backhaul (IAB) to minimize the cost of denser deployments by utilizing the 5G gNodeBs already deployed. IAB would reduce the cost of running fiber to each mmWave access device; however, the technique may introduce more latency.
One of the more impactful observations for 5G transformation is that the networks must morph and scale to optimize resources to support new applications and services. “Scalable” networks was not an explicit goal outlined by the ITU or the Next Generation Mobile Networks (NGMN) Alliance when the 5G standardization process was kicked off by 3GPP, but the implication was clear.
To describe this flexible network topology, the wireless industry defined the term “network slices.” Network slices describes the ability of a service operator to “slice” the network to tailor a unique set of services for users, creating diverse applications and use cases and charging appropriately for the services. With network slices, a company or individual could purchase a service or a set of services to meet specific needs. For example, consider a company that outfits a factory with 1,000 connected sensors. The company may expect to pay less than $40 per month for unlimited data, since those sensors do not transmit or receive the type of data that we consume on our smartphones.
While the 3GPP has, indeed, made great progress toward defining a radio access network capable of achieving these lofty goals, network operators must be able to fairly charge for these services and conserve valuable network resources to sustain a healthy ecosystem and enable all contributors to prosper. To facilitate the creation of network slices, the 3GPP has enhanced the split architecture of the control and user planes to enable separate control and data paths. This is just the foundation. Network slicing also depends on implementing infrastructure elements beyond the physical layer of the protocol stack. Network technologies such as virtual EPC, network function virtualization (NFV), software defined networking (SDN) and mobile edge computing (MEC) are components and services necessary to move network slicing forward. Without these technologies, all data and control traffic must aggregate at the core network, potentially crippling the industry’s ability to meet the goals of data throughput, end to end (E2E) latency and massive connectivity.
Creating the Ecosystem
5G’s success or failure will depend on the creation of an ecosystem. The 5G ecosystem must extend beyond the traditional wireless value chain of usual participants: the service providers, semiconductor companies, infrastructure manufacturers and test and measurement companies, to name a few. Application software and service providers, cloud and cloud infrastructure, vertical integrators, software companies and even car, drone, appliance, medical device and construction manufacturers must be an integral part of the 5G landscape to realize the true economic potential. Creating an ecosystem does not occur overnight. The initial deployments of 5G services to enable the ecosystem to grow and evolve are critical.
A 5G Future
Qualcomm recently commissioned a study by IHS Market to assess the economic impact of 5G. IHS Markit speculates that 5G will become a general purpose technology (GPT), a development so impactful that it becomes a catalyst for socio-economic transformation. For perspective, other examples in our history cited as GPTs include the printing press and electricity. IHS Markit expects 5G to contribute $12.3 trillion—yes, that is a “t”—to the global economy by 2035. IHS Markit is not alone in their prediction, as the world’s economic leaders continue to invest in 5G with a myriad of funding and regulatory support. These global leaders are betting on 5G to catalyze GDP growth and create economic prosperity. While 2035 is 20 years away, the 5G foundation is being laid today. Creating an ecosystem for these applications will take investment, dedication and perseverance and, above all, time. The various ecosystem players must step up to the plate to realize the vast potential that 5G promises.
As noted, initial rollouts are scheduled for next year, and more meaningful deployments will begin in 2019. To realize 5G’s potential, significant innovations must occur in semiconductors and packaging technology, system and network topologies and architectures and, of course, the important verticals that will take advantage of this new network of 5G capabilities and services. It will not be easy, but with the commitment of the industry and the world’s government leadership, 5G has unstoppable momentum.
The industry has made great strides in moving the 5G agenda forward, achieving key milestones in both the standardization process and technology development. Demonstrable commitments from academia, industry and governments worldwide have created forward momentum, yet there remains much to do. 5G below 6 GHz may have a shorter runway to deployment, but mmWave is very important to the overall 5G vision. The next year should provide a better picture of the 5G timelines and potential, and the world will be watching the progress on the challenges outlined in this article. The major players have all anted up, but there are real and hard problems to solve for 5G to live up to its promise. 2018 should be an interesting year and this time next year, a clearer picture of the future should materialize.
The central question is not whether 5G will be impactful, the question is when?