When 4G rolled out, there were two competing communication standards vying for dominance. For a brief period, IEEE 802.16m (WiMAX) and LTE competed to become the next global standard. Things look much better for 5G NR in this respect — the industry is united behind the new standard. Many 5G handsets are already available, and Apple will release its own phones and tablets soon. Networks are already switching on their first 5G cells.
The technical challenges vendors and network operators are experiencing, however, are more daunting than before. They must deploy complex technologies such as millimeter wave communications, adaptive antennas, and virtualization on a vast scale, with 20 or even 30 times as many cells as before. Complex workflows and skill gaps have arisen within their organizations. From a business standpoint, managing CAPEX and OPEX has become more challenging but also more crucial.
3GPP delays standards as buildouts continue
3GPP, the body responsible for defining and publishing the standards 5G is built on, has been forced to delay its work on the Rel-16 Stage 3 and Rel-17 standards by three months due to the global COVID-19 pandemic. The Rel-16 Stage 3 freeze, originally scheduled for April, had been pushed back until June 2020. Similarly, the Rel-17 Stage 3 freeze has been moved to March 2022 – but the timeline may still be subject to change due to the pandemic. That’s a setback for many, but also a potential opportunity for carriers, who could use this reprise to consider adjustments to their 5G strategy as well as component quality, reliability, energy efficiency and rightsizing across all the components in their networks. Because networks with this many sites will demand extreme reliability and exceptionally tight energy management if profitability and sustainability goals are to be met.
A diverse range of service level requirements
One of the hallmarks of the new network will be the way it allocates network resources to its many use cases. Each envisioned 5G service will rely on a different blend of enhanced performance from the network. Key metrics include capacity, data rates, coverage, reliability, and latency. Some current and future use cases are smart homes, the Industrial Internet of Things (IIoT), connected cars, and smart energy grid applications.
These applications can be divided into 3 main categories: enhanced mobile broadband (eMBB), massive machine-type communications (mMTC), and ultra-reliable low-latency communications (URLLC). The first 5G users will be mobile subscribers benefiting from enhanced mobile broadband speeds. The challenge in the beginning will be to provide them adequate 5G coverage, capacity, and speed to keep pace with expectations.
Any way you slice it
After mobile broadband, machine-type communications such as industrial IoT applications will grow rapidly followed by ultra-reliable, low-latency (down to 1ms) services such as self-driving cars. Since these applications all require different service levels, operators need a way to divide up network capacity and bandwidth between them.
Enter slicing. Slicing is based on virtualized network architecture similar to software defined wireless LANs. It enables multiplexing over independent logical networks that all run on the same physical infrastructure. This lets operators allocate resources in real time to meet demand and fulfill very diverse service level requirements. URLLC transmissions can, for instance, be decoded immediately to ensure reliability with eMBB signals processed subsequently.
The business model that slicing enables looks something like this: The infrastructure owner leases out slices of the network to different mobile virtual network operators (MVNO). MVNOs can in turn deploy multiple slices tailored to the requirements of their various applications. Network virtualization was key for many operators in navigating the spiking and changing demand during recent public health lockdowns as well.
Slicing can even be used beyond the radio access network. Companies are already collaborating on approaches that will enable slicing in the 5G core network.
Highband alone is not the answer for 5G
Large amounts of highband spectrum are available, but shorter wavelengths don’t travel as far outdoors and have trouble penetrating buildings. Despite propagation issues, millimeter wave technologies (26-28 GHz) can allow for more densely packed communication links, which can result in more efficient use of spectrum. Attenuation does however mean more cells are required to achieve good coverage. And in an urban situation, that means installing antennas closer to user equipment, at street level, instead of on top of tall buildings or towers. The thermal issues 5G chips create can quickly become noise pollution issues in cities if fans need to operate in close proximity to people.
Where to drop anchor?
Most network operators proceeding with 5G rollouts have chosen, instead of building out a completely new 5G anchor on top of their existing 4G network, to start instead with 5G NSA (non-stand-alone) mode as described in 3GPP Release 15. In contrast to a stand-alone (SA) 5G network, NSA 5G retains the existing 4G LTE anchor, usually in the sub-2.7 GHz range, and builds out 5G on top of it.
This is allowing communications service providers to launch 5G even if they don’t have a lowband 5G anchor in place yet. In 5G NSA mode, user equipment that offers dual connectivity can send and receive data using 4G and 5G simultaneously. The network is transmitting via 4G and 5G at the same time, with the device aggregating the data to provide much faster connectivity than possible with just one of the technologies. Multi-gigabit connection speeds have been measured in real-world conditions.
5G NSA is an outstanding way for communications service providers to start offering 5G immediately since compatible user equipment is already entering the market. But it essentially means installing additional cells on top of the existing 4G network. Since the lowband anchor network does not receive any efficiency improvements, this means increased overall power consumption and greenhouse emissions for 5G. Because no matter how efficient the new 5G equipment is, it’s on top of what is already installed.
Different wavelengths for different purposes
The higher speeds 5G specifies will require new bands such as mmWave (26–28GHz) and wider bandwidths, for instance 100MHz. With 3G and 4G, lowband spectrum ensured sufficient coverage at frequencies below 1GHz with bands for capacity residing above 2GHz. This enabled a wide-scale rollout with acceptable in-building coverage as well as capacity in denser areas. mmWave technology, although able to supply all the bandwidth needed for the high speeds envisioned by 5G, offers poor range and penetration into buildings. So, it’s not the answer for coverage, and coverage is always key in mobile networks. Many of the lower-band frequencies that a 5G RAN could use are however already in use for 4G LTE or even 2G/3G.
Betting the farm on 5G
In areas where 4G is mature and established, CSPs can begin refarming spectrum that is currently still in use for 2G and 3G. Subscribers receive a grace period within which legacy devices will still function. Once that period has passed, the spectrum is available to deliver the lowband coverage and capacity 5G requires. The 2100MHz and 2600MHz 3G bands are also attractive candidates in many locations around the world. 700 MHz, however, would need to be retained for 4G. 1800 MHz could be switched to 5G once most traffic has transitioned from 4G. Of course, spectrum that has been tied to a specific technology by regulators may make refarming spectrum more difficult in some regions.
DSS — our hero, or too little, too late?
How can networks successfully launch 5G if they don’t have access to sufficient lowband spectrum? 5G NSA is certainly a good approach to providing at least some 5G right away, for instance in urban areas. But it can’t deliver the coverage and performance that will be necessary over the long term or in rural areas.
A new technology called dynamic spectrum sharing (DSS) aims to solve this problem. DSS allows 5G New Radio and LTE to transmit over the same band concurrently. This is accomplished with sophisticated software scheduling. But at least one major equipment supplier has yet to fully deliver on its DSS promises, making it difficult for some networks to go forward with DSS.
The strategy for networks who are deploying DSS will be to simultaneously build out their mmWave cells to augment lowband DSS coverage. And DSS remains very attractive because it presents an alternative to refarming, which necessitates the lengthy process of removing legacy subscribers.
Here again, it may prove difficult to reduce power consumption with this type of topology since it lets networks add additional equipment without decommissioning older generations. 5G allows for far greater energy efficiency because 5G equipment can shut itself down much faster and more frequently between transmissions than 4G equipment can. Until a considerable amount of network traffic is running over 5G, however, this advantage may not deliver meaningful overall power consumption and emissions reductions.
Micro-infrastructure promises densification
Small cell options don’t require erecting new 5G towers. In urban environments, large new tower sites may not even be possible when considering zoning and space requirements, not to mention power consumption and the difficulty of, for instance, incorporating renewable energy sources such as wind or solar. Small cells, however, offer an excellent option for increasing density in heterogeneous 5G networks (hetnet).
Microcells are lower power tower sites that cover a smaller geographic area than a standard tower. Picocells offer less coverage (a single building or place of business), but in a compact form factor not much larger than a ream of printer paper. Femtocells are even smaller yet. A femtocell connects to the 5G core through a wired broadband connection and can support between 4 and 8 active mobile devices. The advantage here is that femtocells don’t even need to be powered by the network itself: they can be delivered to businesses who are then responsible for providing a broadband connection and electricity.
Upgrading mobile fronthaul links
Another way many operators have been investing in the lead-up to 5G is by laying fiber to existing cell sites. 5G requires deep integration with very fast wireline connectivity to enable its virtualized RAN architecture. With up to 1,000 times more data flowing for a given area and a stated goal of 1ms latency, existing copper-based time division multiplexing (TDM) circuits connected to the mobile backhaul network will no longer be up to the task.
Some providers have been steadily laying endless lengths of fiber in preparation. Others plan to augment their fronthaul network with directional radio links, which can be just as fast as fiber, but once again will also require power, cooling, and components.
The next wave of services
4G has demonstrated that high-speed mobile broadband can enable new applications that were neither planned nor anticipated. The ubiquitous bandwidth and low latency promised by 5G is poised to usher in its own host of new services. This is why CSPs are investing — the business case revolves around creating new ways to generate revenue with the mobile network. Even governments have recognized the importance of 5G for the economy. In the US, for instance, nearly one billion dollars has been earmarked for 5G research.
A chance to future-proof 5G networks
The unique challenges outlined above will continue to inform 5G network planning for many years. Once 5G has been rolled out on a broader scale to mobile broadband users, the next wave of services will be on the horizon. Massive internet of things and ultra-low-latency applications will require ongoing CAPEX. How much investment is needed will depend in part on the choices networks make now. Will power supplies be dimensioned to meet future needs? Will components be suitable for the compact dimensions required?
The DSS delays are a clear demonstration of how crucial suppliers are to communications service providers right now. With every network doing its best to start offering 5G as soon as possible, technical problems arising from compatibility issues between any number of critical components could cause delays that translate to a strategic disadvantage. In many cases, getting component and module manufacturers involved as early as possible can reduce costs, provide faster turnarounds, and result in more future-proof networks.
The supply chain as a strategic asset
A procurement strategy that focuses exclusively on costs will fail for 5G. Because it cannot adequately prepare for a future that is approaching very rapidly. The challenge is not just minimizing CAPEX but rolling out a much larger network that nevertheless requires as few repairs as possible, minimizes OPEX, and reduces greenhouse gas emissions. Consequently, network operators are being forced to rely on their vendors more across every aspect of the network from base station power electronics to core network data center cooling.
Suppliers that can offer complete solutions can reduce supply chain complexity and fill in skill gaps. In this way, the supply chain and vendor relationships become strategic assets that allow operators to achieve their missions. In a best-case scenario, vendors will also share the operator’s vision of more efficient and more reliable systems—and become true partners for 5G buildouts.