Cybersecurity & Standards, Free Expression
Techsplanations: Part 9, What’s the Deal With 5G?
In the previous post in our Techsplanation series, we talked about core networks’ role as part of larger mobile communications technologies. In this post, we finally discuss how the combination of the technologies we’ve outlined prior can and may work together as part of 5G networks. As before, please refer to this glossary for quick reference to some of the key terms and concepts (in bold).
Loyal Techsplanations readers: you have read many pages of information about networks, spectrum, and way too many acronyms, all leading up to this: a post about 5G. If you haven’t read the previous posts, be warned: we are about to combine all the concepts and technology in the series in this one post, so if you aren’t familiar with things like millimeter wave bands, NFV, MEC, or beamforming, you might want to start at the beginning.
All caught up? Here we go.
You’ve all heard about 5G, but what is it? 5G is the nickname for the next (fifth) generation of mobile networking technology. The capital-G generations are defined by differing sets of standards regarding elements of the network, such as the frequencies used, the data speeds provided, and the protocols for connecting, transmitting, and coordinating information flows across the network. You may have seen the 4G symbol displayed in the mobile connectivity section of your mobile device’s screen. This indicates that the device is connected to a Fourth Generation network, or a network supporting 4G standards. If you’ve seen anything about “10G” don’t be fooled, that’s just the cable industry trying to get in on the hype action.
In fact, many networks claiming 4G status don’t actually meet the requisite standards for “true 4G” network speeds. This has led to a confusing array of network tiers (4G, LTE, LTE-A) with differing maximum data speeds. Long-Term Evolution (LTE) refers to networks “evolving” from older, circuit-switching technology to newer, packet-switching technologies, which allows them to carry more information, faster. According to the 3rd Generation Partnership Project (3GPP), a standard-setting body for network specifications, LTE-Advanced (LTE-A) is the first “true 4G” technology.
This means that even though your device may display a 4G or LTE symbol, the network may not actually meet the official 4G specs. This trend is sure to continue (3GPP already designates “LTE-Advanced Pro” as the next stepping stone) as we move toward the fifth generation standards. Although some carriers have deployed some 5G nodes in a few locations, consistent access to “true 5G,” will not be a reality for most of us for a few more years. 3GPP is still working on defining the standards, and mobile carriers will spend years upgrading their networks to achieve them. Similarly, device manufacturers have only just begun to include the necessary 5G hardware (antennas and chipsets) in a limited set of mobile phones and tablets, so we are really waiting on both the chicken and the egg.
Why? What’s so different about 5G that will take so long to roll out?
There is (at least) one major aspect of 5G networks that will take some time to complete: small-cell deployment. If you recall from the previous posts, the ultra-high speed wireless capabilities of 5G networks will depend on the use of the “millimeter wave” bands of the spectrum. This portion of the spectrum can support significantly higher data speeds, but only over short, unobstructed distances—basically you’ll need to be very close to a base station, and preferably have a clear line of sight to it, to enjoy the potential of 5G’s enhanced speeds. This means that carriers will need to install many, many more base stations to provide consistent coverage with mm-wave cells, and installing that many more pieces of equipment takes time.
On top of installing all these small-cell base stations, carriers must also run fiber optic cables connecting those stations to each other and to the core network. For some locations, this might only require pulling a new cable through an existing pipe, but for others it may be necessary to dig trenches, install new pipes, etc. Even for those places with existing pipes, it will be necessary to add more access points and utility boxes to bring fiber up to street level. Again, lots of labor-intensive work. And remember mobile edge computing (MEC) from the previous post? Deploying all of that distributed computing hardware will take time, too.
Finally, as we mentioned earlier, the standards and protocols for 5G have not been entirely defined. That doesn’t mean that carriers and manufacturers must wait to start building for 5G, but without agreed upon standards they will not be able to build things that need to work across multiple carriers’ networks or need to be compatible with different hardware models. So we are unlikely to see too much development of 5G applications and hardware until the standards are settled.
So, like, how many more base stations will there be?
Good question, and there may not be a real answer for that yet, but the transmission limitations for mm-wave cells likely mean a couple of things.
One, these cells are unlikely to ever be deployed in rural locations—there just aren’t enough subscribers to justify it and it would take way too many cells to provide much coverage. It could be that small towns will eventually get some mm-wave cells, and if there is already fiber near your house, you might be able to get a “fixed wireless” installation that uses mm-wave, but urban areas are more likely to see mm-wave deployments in the near term.
Two, mm-wave cells also face coverage area problems in urban locations—these shorter wavelengths don’t travel well through physical obstructions like trees and buildings. (See part 6 for more about spectral qualities.) So achieving anything like blanket coverage of a city with mm-wave service will require a massive amount of cells, perhaps one on every corner plus more inside buildings.
But mm-wave isn’t the only part of the spectrum that 5G cells will use. In fact, many will use the same (or similar) parts of the spectrum that carriers currently use for 4G and LTE. During the transition to 5G (which could take many years) wireless carriers and devices will use both 4G and 5G technologies. For example, 5G-enabled phones will be able to connect to older networks (4G, LTE) when no 5G is available, and may even make 4G and 5G connections simultaneously.
For a while, at least, it may be difficult to know what kinds of connections your devices have with the cellular network and what parts of the spectrum they use. Carriers will probably deploy both large and small cells operating under 5G protocols. The large cells will use mid-band spectrum (similar or slightly faster than current LTE speeds) to assure consistent, wide-area coverage, while small cells will use high-band (mm-wave) spectrum to provide ultra-high speeds and/or enough bandwidth to support a massive number of connections (like at crowded events). With MIMO (multiple-input, multiple-output) systems, devices may be able to have several simultaneous connections, even with a single cell. This means that even if mm-wave coverage is spotty, you might still experience improved speeds because your data will be transmitted over more than one simultaneous connection, potentially multiplying your overall download/upload speed.
I see. So unless I’m in a city, I may not experience the super fast 5G speeds, and even that might take a while?
Probably. Carriers and the FCC are currently trying to speed things up (this has been and continues to be contentious), but even with an accelerated deployment schedule, it could be years before we see comprehensive mm-wave coverage—and even then, probably only in cities. (And maybe along major highways.)
Well, that’s the kind of hype deflation I came here for I guess. So, if it’s going to be a while before I enjoy mm-wave speeds, what else is there to expect from 5G networks?
Although faster downloads are, for consumers, likely the first noticeable change in 5G networks, full-featured 5G networks are expected to offer many more capabilities. We’ll discuss some of those below, but the potential uses of these capabilities are still in development. Maybe you (after reading this series) will be inspired to design the 5G-supported application that changes all of our lives! To put that in context, it has been said that 2G brought us text messaging, 3G brought the beginning of mobile data and 4G led the way to ubiquitous video streaming (and other data-intensive applications). At the moment, we don’t know what 5G’s legacy will be, but we do know what kinds of capabilities 5G networks might provide.
Ok, I’ll bite. What kinds of capabilities will 5G networks provide?
Fifth generation networks can be thought of in terms of a few key attributes: high speed and capacity, low latency, and the ability to offer customized network configurations. These attributes will allow at least three broad categories of network uses: enhanced mobile broadband (eMBB), massive, machine-type communications (mMTC), and ultra-reliable, low-latency communications (the acronym is just as awkward as the phrase). We will discuss some of the particular use cases currently hyped for 5G networks, but first we will quickly refresh your memory on some of the network technologies involved. (We discussed each of these in more detail in post 6, post 7, and post 8.)
High-capacity small cells – We already covered this one, but these high-speed, high-capacity cells will rely on mid- and high-band spectrum, MIMO, and those time-, frequency-, code-, and space-division technologies (like beamforming) to support high-bandwidth wireless connections for an increasing population of mobile devices. These devices will, of course, include smartphones and tablets, but also new kinds of devices that 5G networks can support, such as augmented-reality and virtual-reality devices.
NFV-via-SDN – I know. 5G (and telecommunications in general) has a serious alphabet soup problem, but what can you do? These particular letter combos (Network Function Virtualization and Software-Defined Networking) are already in use in current core networks, although NFV is expected to play an even greater role in the future. In addition to helping networks adjust more quickly to meet changing conditions, NFV-via-SDN could enable networks to provide services that networks have not provided before. More specifically, the high level of orchestration possible through NFV-via-SDN will enable networks to provide guaranteed levels of service across a suite of network features, such as speed, latency, and reliability. One way of providing these service levels is called “network slicing,” which we’ll dig into later in the post.
Mobile Edge Computing (MEC) – Yet another ingredient in the alphabet soup of 5G. MEC is basically cloud computing, but made of many, many small computing centers distributed throughout the network rather than centralized in one location. MEC offers at least two interesting potential capabilities for networks. First, by placing computing power much closer to end users (say, one wireless hop compared to a wireless hop + multiple wired hops) MEC can deliver much shorter round-trip times for processed data. This low latency (potentially as low as 10 ms) capability is expected to play a crucial role in certain applications such as controlling the flow of self-driving vehicles through a city. Second, having readily accessible computing power close to end users could allow the network to do more of the data processing that our devices (or centralized cloud data centers) currently do. This could lead to smaller, lighter, and cheaper devices that depend on the network to perform much of their processing.
Narrow-band Internet of Things (NB-IOT) – We talked about several methods of enhancing efficiency for spectrum usage in part 7, but we didn’t address this aspect yet. NB-IoT is a Low Power, Wide Area Network (LPWAN) technology, but is also a kind of frequency division. Basically, carriers will use tiny chunks of spectrum to create very low bandwidth channels (where a cellular channel might be about 5 MHz, a narrow band is only 1/25th of that). These channels can be used to communicate with things like sensors that don’t send or receive much information and operate at very low power (which equates to long battery life). All this adds up to a relatively low-cost way to deploy and connect massive arrays of long-lasting sensors.
The anticipated uses of these connected sensors includes everything from utility meters to garbage cans to shipping boxes. Narrow-band channels will allow many, many, many more connections to the network without cutting into the bandwidth used for other devices like phones and tablets, so it appears the Internet of Things may experience rapid growth in the coming years.
So it sounds like 5G will mean significant changes in both radio access networks and core networks, but you’ve already told me not to get my hopes too high for high-speed data. Do all of these changes add up to something else?
Hey! You’ve really been paying attention! That’s right, higher data speeds are definitely part of 5G, but not everything. Perhaps the best way to sum up the (anticipated) impact of 5G is to say that 5G networks will be able to provide exactly the kind of network services demanded by a wide variety of technologies. Where previous network generations were designed to meet the needs of a particular use case (voice telephony and, later, internet access), 5G is designed to provide custom network services for whatever use cases we come up with.
I think I see where this is going. Network operators will somehow combine all those capabilities and acronyms to provide custom services. Is that right? And how will that work?
That’s right! You are definitely going to be ready to impress / bore your friends with your knowledge of network architecture after this. The way carriers will (probably) offer customized network services is through something called “network slicing.”
Network slicing?
Network slicing is essentially a way of creating multiple virtual networks within a single physical network infrastructure, with the option to calibrate each virtual network for a specific purpose. For example, one network slice could carry traditional mobile broadband traffic, providing high-speed data transmission, while another provides ultra-reliable, low-latency transmissions between industrial IoT sensors, such as those on autonomous vehicles, and MEC nodes, while another virtual slice of the network may be optimized for other machine-to-machine communications, such as industrial control systems.
There is still much work to be done to make these ideas real. And although network operators are working on their own infrastructure, they also need the cooperation of online services, device and equipment manufacturers, local governments and municipalities, and other network operators to fully realize the potential uses of 5G networks. Fiber, antennas, edge servers, sensors, and software must all be deployed in a massively distributed system to bring most of the hyped uses of 5G to fruition.
Although these use cases are still somewhat speculative, the potential to reframe the way we consume network-based capabilities exists. The transition to 5G also presents an opportunity for network operators to change the way they monetize networks. Instead of a single, general purpose subscription for mobile voice and data, access providers might like to use slicing to offer a suite of different services, each tailored to specific purposes.
Many uses of 5G networks, and the new billing models that may come with them, depend on network operators’ ability to provide different kinds of service for various applications. But treating network traffic differently runs the risk of violating net neutrality principles. In the strictest definition of net neutrality — that every data packet should be treated equally and identically — network slicing might be unacceptable. But a more nuanced perspective on net neutrality allows for differential treatment, so long as it is not discriminatory in ways that affect the internet user’s Quality of Experience or competition among edge providers.
In an ideal world, end users can enjoy the capabilities offered by 5G technology without giving up the freedom to connect to and use all of the (legal) parts of the web and online services of their choice. At the same time, providers of new products and services should have equal opportunities to tap into the potential of 5G networks without facing discriminatory treatment from network operators and their vertical affiliates.
Stay tuned for additional posts, addressing 5G’s potential impacts on the open internet or even more. And thanks for reading!