While 5G remains an imprecise term today, key objectives for the development of the advances required have become clear. These are as follows:
- Enhanced throughput – As is the case with Wi-Fi, major advances in cellular are first and foremost defined by new upper-bound throughput numbers. The magic number here for 5G is in fact a floor of 1 Gbps, with numbers as high as 10 Gbps mentioned by some. However, and again as is the case with Wi-Fi, it’s important to think more in terms of overall individual-cell and system-wide capacity. We believe, then, that per-user throughput of 50 Mbps is a more reasonable – but clearly still remarkable – working assumption, with up to 300 Mbps peak throughput realized in some deployments over the next five years. The possibility of reaching higher throughput than that exceeds our planning horizon, but such is, well, possible.
- Reduced latency – Perhaps even more important than throughput, though, is a reduction in the round-trip time for each packet. Reducing latency is important for voice, which will most certainly be all-IP in 5G implementations, video, and, again, in improving overall capacity. The over-the-air latency goal for 5G is less than 10ms, with 1ms possible in some defined classes of service.
- Advances in management and OSS – Operators are always seeking to reduce overhead and operating expense, so enhancements to both system management and operational support systems (OSS) yielding improvements in reliability, availability, serviceability, resilience, consistency, analytics capabilities, and operational efficiency, are all expected. The benefits of these will, in most cases, however, be transparent to end-users.
- Increased mobility – Very-high-speed user mobility, to as much as hundreds of kilometers per hour, will be supported, thus serving users on all modes of transportation. Regulatory and situation-dependent restrictions – most notably, on aircraft – however, will still apply.
- Improved security – As security remains the one aspect of IT where no one is ever done, enhancements to encryption, authentication, and privacy are expected. It would not be surprising to see identity management (IDM) solutions along the lines of those now at work in many organizations available from at least a few carriers. Current IDM suppliers as well might be more than mildly interested in extending their capabilities to 5G services purchased by enterprises.
- New spectrum – It is expected that frequencies in the so-called millimeter-wave bands above 30GHz will see service in at least some 5G deployments. Both licensed and unlicensed spectrum at these frequencies is available in many parts of the world. MM wave frequencies are often appropriate to small cells since they require smaller and less obtrusive antennas, and the inherent signal directionality can multiply spectral efficiency. The core disadvantages for MM waves are less applicability to traditional larger cells along with poor object (e.g., buildings) penetration, but such can again be advantages in terms of frequency reuse. Regardless, more spectrum is required given the throughput and capacity objectives that justify 5G development and deployment – present spectral allocations will most certainly not suffice even with the ability to aggregate smaller blocks of spectrum.
- New enabling technologies – We expect to see higher-order MIMO implementations, sometimes described as “massive” with, for example, 16-64 streams, more aggressive modulation and channel coding, improved power-utilization efficiency, and related advances. Small cells will see frequent application, and the days of large cell towers may be numbered in more densely populated areas. Current trends otherwise at work in networks today, include SDN and NFV, will also see application in 5G, with much infrastructure implemented within cloud-based services. 5G will likely require no major advances in chip or manufacturing technologies, and device power consumption will likely benefit from more limited geographic range even as higher clock rates take a small toll here. Still, much work remains in terms of both technical and feasibility analysis as well as cost, but we see no showstoppers on the horizon. There is no danger of producing another WiMAX that offers marketing hype with no clear advantages over the previous generation, and the overall level of technical risk is low. Perhaps the greatest challenge is schedule slip, as the complex nature of the systems engineering that is required needs more time than many expect.
- Universal application support – 5G as a wireline replacement will have to support every class of traffic and every conceivable device, from broadcast-quality video distribution to telemetry, implantable medical devices, augmented and virtual reality, and advanced interactivity and graphics – and not just for gaming. The list also includes connected and autonomous cars, remotely-piloted vehicles (drones), public safety, building and municipal automation/monitoring/control, and disaster relief. including relocatable infrastructure with moving cells and support for dynamic wireless meshing. Also in the mix are robotics and IoT devices tolerant of limited data throughput and highly-variable latency. We expect literally tens of billions of 5G devices to be deployed over the next decade or so, so the scale of both the challenge and the demand is clear.
- Industry growth – Finally, carriers, operators, and equipment vendors of both infrastructure and subscriber devices simply require the deployment of new technologies with quantifiable end-user-visible benefits from time to time in order to continue to grow their businesses. New subscriber units alone cannot accomplish this goal.
In short, 5G is a business opportunity being designed and implemented to provide all of the communication capabilities and performance we expect from a wireline network. Getting to that point, given all of the requirements above, won’t be easy, quick, or inexpensive.