From the factory floor to security, safety, health care, the automotive industry and more, we expect that the combination of 5G, processor and AI capabilities will be embedded
and imbued into everyday machines and devices. This could have implications for the industry that far exceed faster download times on our mobile devices.
At the same time, several challenges remain as telecommunications providers work to establish 5G-enabling networks and infrastructure. According to the IBM Institute for Business Value report, The end of communications services as we know them, when asked about what percentage of services will be uniquely enabled by 5G capabilities versus enhancements of 4G services over two and five years, respondents reported that the vast majority of services would not be uniquely enabled by 5G.
In two years, respondents reported that only 7% of services would be uniquely enabled by 5G capabilities, and 93% would be enhancements of existing 4G services. Even in five years, respondents expect that only 18% of services will be uniquely enabled by 5G. This is a problem because it suggests most people think 5G is only about faster speeds, and not about enterprise transformation.
Organizations remain wary about complexity and costs, as well as external challenges, like supply chain issues and concerns about the impact of 5G radio signals near airports. Yet, the opportunity far outweighs the cost. 5G would grow the portion of the economy driven by digital platforms to more than half of the global economy, and unlock new economic opportunities in immersive entertainment, augmented reality (AR), connected vehicles, Industry 4.0 and more, according to a report by IBM. Navigating the shift to full 5G deployment, while also providing full enterprise 5G coverage for the existing plethora of handsets and network devices, means starting the private 5G journey now by taking some of the following actions.
Target small pilots
One effective strategy is to target small 5G pilots that can be built quickly to demonstrate value and return on investment to enterprises, and then scale up to realize the benefits. Targeting smaller pilots, which can then be scaled up, makes it easier to align with your objectives.
Adopting a Cloud-Out Network Architecture
There is also a cloud angle to the 5G transition as the rollout will require higher spectrum bands that do not travel as far as lower spectrum bands. Much of what counts as 5G technology is virtualized software, not physical boxes, and this software can run pretty much anywhere.
Embracing a hybrid cloud model that gives you a full picture of network performance can help you expand service while minimizing outages and disruption.
Scaling 5G networks which may end with telcos deploying ten times the number of basestations over 4G requires a radical rethink on automation and management.
Many telcos have struggled with automation so far not because they haven't started the journey, but because they've built scripts to drive automation tasks that are fixed to the underlying hardware.
These scripts can be incredibly brittle, meaning that changing a software version or a hardware element in the chain requires a rewrite of the script. Paradoxically, with more scripting, the network becomes more brittle and it takes longer to make changes to the network or adapt to new software or configurations.
There is an alternate approach, which abstracts the hardware from the business logic, uses scripts to drive the intent of the network and lets the automation engine figure out how to translate this business intent to device execution. When tied with AI and dynamic service chaining, the network can adapt to failures, changes and attacks in real-time.
Surface issues that matter via intelligent observability
To help provide a smooth transition to 5G and AI-powered network automation, engineering managers need real-time visibility into network performance across physical, virtual and software-defined infrastructures to spot performance issues early, address them quickly and avoid costly performance issues that can impact business and end users. Advanced application and network performance management have the capability to gather data from multivendor sources, blend it with the network data and then use AI to spot patterns and anomalies automatically.
Network managers face a multitude of challenges such as latency, poor performance and down-time pinned because a business unit made network changes that ignored the constraints posed by geographic distance and bandwidth needs. Without tools specifically designed to monitor, control or secure multi-cloud networking, network managers are often left with a fragmented picture of their network connectivity between multiple clouds.
However, by embracing these tools and fusing AI into their network management, they can get a cohesive picture that will help them reduce disruptions while scaling up new 5G networks more quickly and securely.
The full rollout of 5G breeds higher expectations from consumers for quicker, more expansive coverage, newer and more powerful services and stronger AI. Delivering on these expectations will require an advanced, 21st century connected communications infrastructure. Communications services providers and the organizations they work with will need the efficiency and insight that AI-powered automation can provide to meet the challenge.
Andrew Coward, General Manager of Software Defined Networking, IBM
Photo by ThisisEngineering RAEng on Unsplash