Chris Pearson, President, 5G Americas – November 2022
It has often been said that the Fourth Industrial Revolution (4IR) represents a merging of cyber-physical systems, where the data generated in the cloud and on networks will help to inform and animate objects and things in the real world. It could be the natural evolution of entire economies that have been built on information, as well as electricity and steam before it.
But where many 4IR concepts remain in the realm of localized manufacturing and on the factory floor, modern data and communications networks are much more robust and pervasive, spanning across virtually all industries and reaching into virtually every facet of life. Indeed, the latest Ericsson Mobility Report update shows that mobile data traffic use has now reached 100 exabytes per month, demonstrating the sheer massive quantity of data being consumed. Whether you use the Internet for work, play, travel, or organizing your life, there really is no longer any meaningful escape from being connected to every other person or thing on the planet. We are connected.
Even if you are wandering through the Sahara Desert without a cell phone, there’s a chance that a satellite may be capturing a photo or video of you. So unless you find yourself in the deepest recesses of an underground cave in Borneo with no electronic equipment, you are likely being impacted by the new future fabric of data and communications networks. It is this ubiquity – this near total pervasiveness of the overlay of data on top of everything – that gives such awesome power and responsibility to organizations to use that power to improve (not harm) lives.

In the latest 5G Americas white paper, “Distributed Compute and Communications in 5G” we look at how the growth of cloud computing is becoming ever more distributed and pervasive around the world – and what 5G is doing to ensure people and things are connected when there is no “wire” in the last mile. We also examine how edge cloud computing is transforming 5G wireless networks in terms of extent of distribution and diversity of compute and storage capabilities, enabling management of data from a massive number of wireless devices. Finally, we look at how a Distributed Compute and Communications Fabric (DCC-Fabric) that overlays the world’s networks will need to be created in the future to address tomorrow’s challenges.
When you think about it, this modern age of ubiquitous cloud-based connectivity hasn’t been around very long. Indeed, while the notion of network-based computing dates to the 1960s, many people believe the first use of “cloud computing” in its modern context occurred on August 9, 2006, when then Google CEO Eric Schmidt introduced the term to an industry conference. Some would argue the modern smartphone era kicked off a year later in 2007 – and 4G LTE launching three years later in the US at the end of 2010.
From the outset, cloud computing and wireless communications have been joined at the hip. Mobile wireless communications brings the power of the cloud into the real physical world.
But as real-world data use continues to increase dramatically (doubling every year or so), processing everything at the data center has become challenging. Sometimes it seems, there simply is not always enough infrastructure to move around all the bits of data to the data center quickly enough for many demanding use cases. Today’s cloud and communications systems might not be capable in the future of capturing, transmitting, storing, and analyzing the petabytes of data generated by the soon-to-be trillions of sensors operating 24/7. Indeed, today’s cloud and communications systems were designed to deliver services with less stringent requirements to people over long distances (not devices over short distances) and could need to be bolstered by highly distributed caches of Content Delivery Network (CDN)-hosted content.
Current systems face future challenges to deliver the needed compute for real-time AI inferencing required for the factory of the future, mixed reality (MR), and Extended Reality (XR) with haptic interactions, connected vehicles, assisted living, or merging of physical and digital worlds with 5G and subsequent generations of systems. So cloud resources are moving toward the edge of the network – even all the way down to the device so they will meet performance, cost and/or legal and regulatory requirements.
How do you manage mission-critical performance in an automated factory operation when all the mobile robots need to move synchronously with response times at 10 milliseconds or less? You’re certainly not going to send all that data to a far-away data center for processing. You’re going to do it locally. This is where edge cloud computing is playing a critical role in enabling emerging use cases with extreme service requirements in a variety of sectors. Increasingly, this is being addressed through the emergence of “edge zones” that manage local data flows – depending on the amount of data traffic and number of end points.

A lot of these new changes are being achieved using traditional commercial-off-the-shelf (COTS) hardware that is being augmented by specialized programmable hardware resources to satisfy the strict performance requirements of certain applications and limited energy budgets. So overall, as the commercial applications become more sophisticated and demanding, cloud computing resources themselves are becoming increasingly heterogeneous and complex – and becoming increasingly widely distributed across many smaller data centers.
Let’s noodle on that for a moment: imagine you’re operating a modern factory. You have parts that are running along a conveyor belt, which were shipped this morning into your receiving area. You need to assemble a million widgets from this mish-mash of supplies, get them boxed up, labeled, and ready to ship out. Every single one of these pieces might require a different kind of specialized network use:
- Your receiving area needed to track and manage inventory, sending data to your buyers and receiving tracking information to make sure the supplies arrived okay – which is generally not a very time-sensitive application and text or a spreadsheet works pretty well.
- But then your robots will need to visually process the inventory to make sure nothing is broken. Their cognitive visual AI might need a lot of video data to assess the integrity of the supplies. Is it exactly the right part? Video requires a lot of data – and high-definition video requires even more.
- You need to move your parts to the floor with the robots, which will again need high quality video to navigate the space. This time with very little network latency or else they might bump into something or someone. Mobility AI will need a lot of training data to understand barriers and obstacles.
- On the conveyor belt, your robots have a fixed field of view, but their reaction time is critical. These require intensely low latency network connections.
- Boxing, shipping, inventory management, fraud detection all require different kinds of network connections.
- Additionally, with vendors, suppliers, and contractors all moving into and out of your facility, you must address security for all the sensors, tags, vehicles, and mobile devices they operate within your network.
- Finally, all the factory operations data must then be transmitted to headquarters to integrate with your firm’s cloud-based storage and central data lake for business analysis, billing, etc.
So what happens if your organization and all your business partners are using different hardware, software, and systems? Now you got a challenge to overcome. This where the Distributed Compute and Communications Fabric (DCC-Fabric) comes into the picture in the future.
For all this to work synchronously, a DCC-Fabric consisting of wireless and wireline access to the Internet in distributed schemes can underpin ubiquitous compute and communications needs. It can be an automated lattice of overlay connections on top of the physical topology of physical/wired connections that provides a scalable, robust foundation for communication among entities and network devices. It will have to span, management, orchestration, and control – as well as use extensive intelligence enhancements to safeguard information flow, privacy and security. It will also need to address heterogeneity – or a variable mix – of systems (even special nodes), protocols, services, security and trust mechanisms, technology domains, administrative domains, governance models, ownership, economic models (pricing, accounting, charging, billing), and current or future applications.
Here’s what a DCC-Fabric organization might look like:

But the smart factory floor is just one example. A DCC-Fabric is going to need to be built to encompassing virtually all facets of the modern data network. It will be central not only to Industry 4.0, but also have major impacts for addressing social and economic goals. For instance, at scale, the DCC-Fabric can help address sustainability and climate change.
Through applications like IoT sensor tracking of penguins to dynamically pulling in metadata on crop management efficiency, can help solve some of humanity’s toughest problems. Sensor tracking can address CO2 and other atmospheric emission reduction goals. Deploying network nodes closer to users and caching allows for reduction of overall backbone traffic, which also reduces energy spend. Smaller, distributed data centers may also take advantage of local renewable energy micro-grids.
A DCC-Fabric can also address digital equity and rural/urban access divides through availability of affordable, robust broadband internet service, as well as applications and online content designed to enable and encourage self-sufficiency, participation, and collaboration. It could help address the medical sector, including the need for telemedicine in rural areas, – and also deliver an overall economic boost through sophisticated applications in MR, Extended Reality (XR), metaverse, factory of the future, assisted living, and connected vehicles – among others.
These benefits will not be without their own challenges. Today, creating a DCC-Fabric involve combining many key ingredients for success, including coordination across industry players to create a common set of specifications and APIs to ensure proper interoperability, as well as brokering access (leases) to resources in a fair manner due to the complexity of reserving resources to give users a seamless and intuitive experience. Additional challenges include regulatory compliance, security, open marketplace, access and reliability, cost and sustainability, managing complexity, synchronization, and managing explosive growth.
While the challenges to creating a DCC-Fabric sound enormously high, the benefits appear to far outweigh the obstacles. This is the direction that our networks could be headed towards – one that is interoperable, capable of autonomy, and serves the common needs of humanity. Just like the earliest weavers who learned how to spin thread and turn those threads into cloth, we are at the beginning stages of a new era of interconnectivity in this new era of 5G innovation. The new fabric of data networks is an exciting concept. 5G networks will play a crucial role in this new fabric, ensuring the creation of cyber-physical systems, blending cyberspace with the real world, and advancing the inexorable progress of connectivity.
-Chris