Why software is key to 5G
5G is one of the hottest topics in the telecom industry today. We hear that major carriers are rolling out 5G in select locations across the country and that most carriers plan to have operational 5G networks within the next several years. But the promises of lower latency, higher capacity and greater speed will be put to the test in the customer experience. Will 5G fall short? The answer lies in how carriers are anticipating the scale at which 5G will change their operations, and how they’re preparing to implement 5G.
Traditional networks, like 3G and 4G, were architected around purpose-built, closed hardware (and software), such as the equipment on cell towers that dot the U.S. landscape. This hardware provided the central foundation for a carrier’s backend network infrastructure and was designed to provide voice services and internet access. But the hardware and software design of the past can’t deliver the performance of the future. Those same cell towers, the 4G macro cell sites, for instance, aren’t designed to handle true 5G specifications and consequently won’t be able to provide the benefits of true 5G.
Based on the advertisements and some pundits, one might assume the future of cellular networks is already here. It might even say 5G somewhere on your phone. In actuality, carriers have been rolling out some of the 5G specifications, but they’re still running on a 4G infrastructure. One carrier may call its network the “5G Evolution,” but that network is still running on an existing 4G LTE network, while providing better 4G speeds. Other carriers are deploying what’s called Non-Standalone 5G, which is also a solution running on 4G infrastructure. In all these cases, the potential 5G speed, throughput and bandwidth are all limited by the 4G hardware and network design.
Why it’s different
5G will enable ultra-fast, high-bandwidth content. It’s the stuff of the future: extended reality, autonomous/smart vehicles, smart factories, smart cities, smart home and mission-critical applications like e-health and telesurgery. To power those use cases, 5G requires ultra-reliable and low-latency communications, massive machine-type communication and enhanced mobile broadband—the 5G triangle. These capabilities can only be achieved with Standalone 5G NR (New Radio) networks, running on native 5G infrastructure. Just one example of the kind of infrastructure shift 5G necessitates is device density support (i.e., how many devices can the network handle). The International Mobile Telecommunications-2020 standards for 5G require a minimum connection density of 1 million devices per square kilometer, while the current 4G Low-Power Wide Area standards support 60,680 devices per square kilometer. That sort of infrastructure overhaul is no small task.
So, what does this mean for carriers with respect to implementation and operations?
There are many aspects, attributes and processes to efficiently and effectively implement and operate 5G networks, but one of the key components, if not the most important one, is software.
Software is key
With 5G NR’s wider spectrum at higher frequencies, there are some notable challenges. 5G is not as good at penetrating obstructions like hills and buildings and traveling longer distances. To solve this, 5G networks are implemented with smaller equipment arranged in closer proximity.
In the 4G LTE macro site model, you have giant cell towers relaying signals that look like a directed graph. 5G looks more like a mesh network with many more “cells,” cross-connected for coordination purposes. These smaller cells can be mounted on lampposts, traffic signals, in buildings, pretty much everywhere. There will be a trend to use more commodity type (open) hardware to handle network functions like firewalls, DNS and encryption in a virtualized environment. 5G will rely on a software-based concept of Network Function Virtualization. Handling the more complex topology with simpler, smaller hardware puts the onus on software to effectively run a 5G network.
On the operations side, you have concepts such as network slicing, which enable carriers to build virtual end-to-end network slices or chunks tailored to the specific applications. For instance, you can have a slice for mission-critical IoT applications such as autonomous vehicles or telemedicine that function on ultra-reliable low latency, while there could be another for handling high bandwidth, such as mobile broadband, and another one for manufacturing or retail.
These are some of the many concepts that are part of Software Defined Networking (SDN) and will be heavily employed in implementing and operating 5G. Because SDN is software-based, there are opportunities for equipment vendors, services vendors and the carriers themselves to leverage smart software components, such as machine learning, to build and operate an efficient and powerful network to realize its promise. These require a solid software strategy to implement and operate correctly and efficiently.
What carriers should be doing now
Carriers are already making use of existing 4G infrastructure to provide some of the promises of 5G. They’re mostly doing this as an incremental step in the evolution towards full 5G or Standalone 5G NR. There’s nothing wrong with that, and in fact, it’s logical to put in place the coverage now and migrate later. When it comes time to migrate, again software comes into prominence, especially as carriers evolve from legacy networks by making large densification upgrades on the Radio Access Network.
Preparing for the effective management of a network rollout will be essential to its success. As mentioned earlier, the topology is more complex, thus making the rollout more complex. Take, for example, cell site lifecycle management. It used to be just sectors and back-haul. Now you have front-haul, mid-haul, back-haul, as well as the shift from a directed graph network to a mesh one. All aspects of the cell site lifecycle have become more complicated. These include:
- Site planning (what capacity is needed where).
- Site selection and detailed design (what elements will be placed where, operating at what frequency, over what physical sectors, interconnected via what topology).
- Site development (permitting, construction, provisioning, testing and validation, indicating on-air).
- Network operations and monitoring.
- Physical site maintenance (contracts, notification lists, access controls, inspections, hazard management).
- Compliance verification.
The systems that support these functions need to adapt. Planning and implementing cell site rollouts has historically been a complex affair for carriers. Having the right software strategy that makes each of the areas of cell site management go smoothly is critical. This will emphasize the need for intelligent automation, RPA, distributed ledgers, AI or machine learning as appropriate to not only speed up operations but also provide a better experience for all the “actors” in the lifecycle.
The 5G revolution promises life-changing experiences for all of us. In some ways, the future has already arrived. All one needs to do is look around the house to see the impact. From your TV to your phone, to your printer and Alexa, and the camera on your doorbell, all these consumer electronics are connected. Expand that to include autonomous cars, smart-meters and manufacturing sensors, and you start to get the picture. We live in an increasingly connected world more dependent each day on reliable internet service and bandwidth.
While it’s easy to see the importance of the hardware components of 5G, the underlying invisible foundation on which these experiences are built is software. It’s imperative for all actors in the connectivity supply chain—equipment vendors, carriers and service providers—to place an emphasis on software development and intelligent automation. Without them, it may say 5G, but in the real world, the promises of 5G will not be realized.
Learn more about our work in telecom and media.Tags: 5G, Edge, Software, Telecom