The forthcoming generation is not a forklift replacement so much as an extension and evolution of existing 4G mobile transport infrastructure – but “let’s wait and see” is not an option for network operators.
Looking closely at their infrastructure through a 5G lens, go-ahead mobile operators see opportunities to ensure that all upgrades and extensions will be steps in the right direction – toward the 5G future. Jon Baldry, metro marketing director at Infinera, explains.
With the first commercial 5G rollouts now announced for 2018, it is highly likely that the first 5G handsets will be announced at next year’s Mobile World Congress, initially at a premium price for people who demand the ultimate despite minimal 5G network availability. It’s human nature – like desiring a Ferrari in a world with a blanket speed limit of 70 miles per hour.
This will make headlines for sure, but the real news is not so obvious. With the transition from 2G to 3G to 4G, the public has gotten used to the idea that a new standard means a whole new network, with providers competing to be the first to roll it out. Operators are still vying to be the first to market with 5G, but this time we are not replacing the 4G network, we are extending its reach into smaller 5G cells and evolving toward new 5G standards. From the outset, 4G was not set in stone: ever since its launch there has been a continuing series of further 4G releases to add new functionality, many of them geared toward supporting new 5G networks.
Why 5G?
The real driver for 5G is not a single set of applications, but a broad group that is typically clustered around one of three centers of gravity: enhanced broadband, massive machine-type communications and ultra-reliable, low-latency communications.
Figure 1: The 5G Services Ecosystem source ITU-T
Enhanced broadband naturally extends the capabilities of 4G bandwidth per user to enable 5G to challenge in the residential and business broadband markets. 5G will deliver speeds at least ten times faster than 4G to enable cloud storage of high-definition (HD) video clips and support for 4K video.
Massive machine-type communications is geared toward the Internet of Things (IoT), with up to a million connections per square kilometer or 100 devices per room, whereas 4G now manages a few thousand per cell. Although initial IoT deployments have centered on a population of very simple devices such as traffic sensors and smart meters sending and receiving relatively tiny pulses of data, these are just the ripples preceding a potential tsunami.
Connecting surveillance cameras to the system will add a lot of traffic, but it is forthcoming ultra-reliable and low-latency communications applications such as driverless cars, industrial control and telemedicine that will really pile on the pressure. These applications require secure communications that never fail, with network latency dropped by a factor of 10 from 4G standards, to an impressive 1 millisecond, to give undetectable response times.
Many of the envisaged 5G services will use a blend of these capabilities. For example, virtual reality (VR) will require both high capacity and ultra-reliability with low latency. VR is more than just a game: it has the potential to transform education, training, virtual design and healthcare. If a surgeon is to diagnose accurately, or even perform a remote operation, the resolution of the virtual reality image must be close to the resolution of a human retina. This requires at least 300 megabits per second, almost 60 times higher than current HD video, with undetectable latency and of course ultra-reliability.
These are the sort of facts, figures and exciting applications that make the headlines. The real work, however, is to provide a whole network that can support such service levels. What does this mean for the mobile provider who already has a huge investment in 4G infrastructure?
How to Get There
5G achieves its massive bandwidth by operating on higher frequency bands, in the millimeter wave spectrum. At these frequencies, signals do not travel as far and are more readily obstructed by walls, obstacles, rain or mist, requiring clear line-of-sight access.
A slow download to a smartphone or a break in a phone conversation is annoying but seldom disastrous – and earlier technologies were no better. Extreme reliability is, however, essential for driverless cars or other critical 5G services. We cannot afford any blind spots where a building shadows the signal.
Full availability means that many more smaller cells must be added to the network. Existing 4G access must be extended like capillaries in a fine network of small cells feeding back to existing transport arteries. This requires a huge investment, partly compensated by the fact that 5G antennae can be much smaller and use less power. They will also conserve power by focusing signals more accurately rather than beaming equally in all directions at once.
To support this more dynamic cell behavior, we need greater intelligence toward the edge. As well as using multiple antennae to aim signals more efficiently, 5G will also recognize the type of signals being sent and reduce power when less is needed. Having a host of small cells in close proximity also enables coordinated multipoint (CoMP), a technique in which nearby base stations respond simultaneously and cooperate to improve quality of service.
While the new radio access network does its best to minimize latency, this comes to nothing if some signals have to travel all the way to and from a distant data center. So another trend will be for mobile edge computing (MEC), in which caching, compute power and critical applications will be pushed closer to the network edge to reduce latency and congestion in the transport network and optimize quality of service.
Fiber-deep Challenges
Existing cellular networks rely heavily on fiber optic links to connect cell towers to the core network. Although high-speed wireless can bridge the gap when time or cost makes it impossible to lay fiber, the only technology to consistently support 5G’s surge in demand and quality of service will be fiber. Each cell of a capillary 5G network is far smaller than a typical 4G cell, but there are so many of them and the applications so demanding that the total bandwidth demand in the transport network will be massive.
Figure 2: 4G and 5G Cell Coverage
So it is necessary to extend fiber as close as possible to the small cells in order to meet this demand. This “fiber-deep” evolution will not be achieved by simply multiplying existing fiber equipment and building it out into the metro space as needed – that would be a colossally expensive operation both in terms of real estate and equipment costs. Instead there will be a need to install many more compact and power-efficient network nodes wherever they can be economically accommodated. This could include remote telecom huts, street cabinets, cupboards or cell sites – locations quite unsuitable for housing racks of equipment that is optimized for a controlled telco environment.
Selecting suitable equipment will no longer be a simple matter of asking a preferred supplier to meet the required performance levels. It will be necessary to look much more closely at the specifications to see if devices are sufficiently rugged, compact and power-efficient to survive where space and power supply are limited and temperature and humidity levels more extreme. With a massive increase in the amount of fiber installations, commissioning and operating expenses will also soar, unless extra care is taken to choose the most compact, reliable and easy-to-maintain optical equipment.
Leading optical equipment suppliers are well aware of these challenges and are developing solutions more suitable for fiber-deep networking. The latest access-optimized units can deliver 100 gigabits per second (Gb/s) at a mere 20 watts, packing over 400 Gb/s into one standard rack unit – about eight times the density of previous-generation equipment.
What’s more, the industry has been working to bring the International Telecommunications Union’s (ITU) vision of autotuneable wavelength-division multiplexing – passive optical network (WDM-PON) optics up to the performance levels required to support the reach and capacity requirements of 5G networks. This eases the pressures of commissioning and maintaining extensive dense wavelength-division multiplexing (DWDM) optical networks by replacing the technicians’ burden of determining and adjusting wavelengths at every installation. Autotuneable technology will automatically select the correct wavelength without any configuration by the remote field engineer, enabling them to treat DWDM installations with the same simplicity as grey optics.
Pressure on the Transport Network
This far denser 5G access environment, even with greater intelligence located toward the edge, will put heavy pressure on the upstream infrastructure. In between times of change, buying patterns tend to stabilize toward the convenience of familiar, single-vendor provisioning. With the shift to 5G we are already seeing greater competitive pressure between mobile operators, and between wholesale operators hosting 5G transport services. This is forcing buyers to demand higher performance, greater efficiency and more demanding specifications – driving a shift toward more aggregated best-of-breed solutions.
Higher performance is not all that is needed – there are other significant changes taking place as 4G networks evolve toward 5G. Data center technology, such as spine-and-leaf switching and network slicing, will increasingly migrate to the transport network to provide the flexibility to support more distributed intelligence and the need for MEC. While 4G started with high-performance dumb pipes connecting cell towers to the core, we are now evolving toward a more flexible software-defined transport architecture.
As well as greater capacity, there are other demands that will not be met by many existing optical solutions. Among the refinements required for 5G, carrier aggregation enables the use of several different carriers in the same frequency bands to increase data throughput, rather as CoMP (described above) makes use of neighboring cells. These solutions require new levels of synchronization precision, as well as low latency. Mobile operators now buying equipment need to look closely at the specifications to ensure that they are not investing in systems that will become obsolete as 5G rolls out. There are already some nominally 4G mobile transport networks that meet the demanding 5G synchronization and latency specifications.
Conclusion
5G-readiness is an ongoing development, and we can expect more early announcements of 5G services on the basis that they meet 5G speeds or other criteria without providing full 5G mobile service. Like owning a Ferrari, it’s a combination of marketing hype and status. Providers and nations are understandably keen to demonstrate 5G way ahead of the timelines favored by the 3GPP standards body.
Major sporting events, with their massive global television coverage, offer a stunning opportunity for operators to showcase their 5G capabilities. The 2012 London Olympics were the first “smartphone Olympics,” where live spectators could simultaneously view the games up close on their handsets. The 2020 Summer Olympics in Tokyo and the 2022 Winter Olympics in Beijing will vie with each other to highlight the way these nations are driving mobile 5G, as Europe once drove 3G and North America drove 4G. Europe and North America are also looking to showcase 5G, as in Elisa’s recent announcement of what is claimed to be the world’s first commercial 5G service in Finland. By 2022, there could be a significant number of Beijing Winter Olympics spectators using 5G VR devices to spectacular effect.
Meanwhile, mobile operators need to work steadily toward these capabilities with 5G-ready mobile transport that will optimize 4G networks today and provide the high performance required for full 5G in the future. Operators can avoid investing in soon-to-be-obsolete mobile transport technology by seeking advice from experts at the leading edge of optical network equipment and design.
