In a series of memos written in 1962, J.C.R. Licklider, a director at the U.S. Department of Defense Advanced Research Projects Agency, proposed an “intergalactic computer network” in which computers around the world would be interconnected to provide quick access to data and programs from anywhere.5
Only a few years later, in 1965, Leonard Kleinrock, Lawrence Roberts, and Thomas Merrill used telephone lines to connect four computers together, effectively creating the first wide area network as well as the beginnings of the Internet.6
More than 50 years later, Licklider’s original vision remains the same, as the network continues to connect global consumers of information and services to applications and data sources.
Everything else, of course, has changed.
Today’s world, driven by the exponential growth in technology performance, has become increasingly connected, digitized, distributed, and diverse. With every “thing” possessing the power to process data, compute models are poised to become dramatically more distributed and networked. And as devices and users are added to it, the value and importance of the network as measured by Metcalfe’s law continues to grow exponentially.
Digital business continues to fuel networked innovations. IDC estimates there will be 48.9 billion connected devices in use around the world by 2023,7 and the 2018 Cisco Complete VNI Forecast predicts that the average amount of data consumed across a network will be almost 60 GB per personal computer per month.3
It’s not surprising that, given this relentless growth, we find that IT teams are reaching a point where the sheer scale and complexity of networks is exceeding their ability to manage and secure them effectively. What’s needed now are new systems that combine technologies such as machine learning, machine reasoning, and automation to simplify operations and augment human decision making.
We are now on the threshold of a new era of networking, when IT can break with the traditional ways of building and operating networks and embrace a future powered by technologies that can solve these challenges in vastly new ways.
Before we examine the emerging network technology, operations, and talent trends that form the basis for this new network, let’s briefly consider the global business and technology trends driving its evolution.
Applications and data, of course, are at the heart of digital business, and how applications are developed, hosted, and consumed is constantly changing to meet new business needs.
Here are some of the ways that applications are evolving, and in some ways, reshaping the network:
Apps and data are leaving the premises: Applications and data are being modularized into microservices and moved to multiple public clouds. In some cases, they are also being distributed to the network edge. And they are increasingly being consumed from multiple software-as-a-service (SaaS) providers.
Apps are modular and distributed across environments: Monolithic applications are in many cases dissolving into interconnected microservices that are delivered via a variety of virtual and physical workloads, including containers, across the entire enterprise.
Apps are being built continuously and rapidly: For applications developed and hosted on-premises, IT has to accelerate its own infrastructure service creation and delivery to meet the needs of applications and users, all while containing operational costs.
Apps are migrating from physical to virtual to containers to serverless: The rise of containers is exposing application design and deployment paradigms to a much more massive disruption, namely serverless architectures, which is forcing organizations to reexamine how applications are built, the role of infrastructure, and the design of operational processes.
With applications and microservices popping up across all domains, the network should be seen more as a growing set of interconnected “nerve clusters” situated where the data is, which could be anywhere along the edge-cloud continuum. The new network needs to be able to securely connect within and between these interconnected “nerve clusters” as well as to fundamentally understand how these new application models work and to dynamically extend application policies across the network to wherever the applications are hosted.
The exploding use of IoT devices, apps, and accompanying data is driving the creation of new distributed compute models that consist of exponentially greater levels of scale and complexity. According to the Cisco “VNI Forecast Highlights Tool,” machine-to-machine (M2M) devices will account for 51% (14.6 billion) of all networked devices around the world by 2022.12
In addition to providing connectivity and security for an incredibly diverse range of IoT devices, network administrators will need to devise scalable and efficient ways of automatically identifying, classifying, and applying policies and monitoring them to ensure proper functionality without impacting or compromising other services running on the network.
The emergence of AI-powered apps for both business and consumer use is leading to a whole new world of connected, intelligent, and automated devices that are being deployed everywhere.
To unlock the full potential of AI in business, more computational processing and decision making have to be done closer to the edge. Depending on performance, capacity, privacy, and even cost considerations, the placement of AI processing and data will range from the cloud to on-premises data centers to the edge of the network.
According to the Cisco “VNI Forecast Highlights Tool,” global business mobile data traffic will grow sixfold from 2017 to 2022 at an annual growth rate of 42%.12 Business mobile users will continue to expect immediate and high-performance connectivity anywhere, anytime, and on any device over Wi-Fi and public 4G and 5G networks. At the same time, wireless IoT devices will become increasingly ubiquitous in all aspects of our lives.
Employees accessing cloud applications from corporate and private devices when off network is creating a lack of visibility and control that network and security administrators haven’t faced. And a wave of IoT devices will add to wireless networking requirements in terms of scale, different traffic patterns, and security.
Cybersecurity threats are becoming increasingly sophisticated and dangerous across a broader attack surface that is no longer contained within well-defined and defended perimeters. In particular, as workloads move off-premises, there is a danger that IT will lose visibility.
While the network will continue to be a powerful ally in identifying and containing threats, network and security operations need to share data and integrate tools and workflows to best combat the continued rise in number and sophistication of attacks. In addition, the network can extend the reach of IT into cloud environments to help protect applications and data even when not directly under their control.
Increasing video usage and the emergence of virtual reality (VR) and augmented reality (AR) for improved collaboration, training, productivity, and remote working experiences will place ever greater demands on an organization’s network.
The network will need to provide the end-to-end bandwidth and low-latency communications and dynamic performance controls required to enable such immersive experiences.
Not only is this dynamic technology landscape a reality for all organizations and their customers, but it’s also the engine for the digital economy. It’s no surprise then that IT is feeling the pressure to address all these trends with the right networking technology strategies, operations models, and talent.
For an organization to flourish in the digital economy, the network needs to be able to adapt quickly to changing business requirements. The network needs to support an increasingly diverse and fast-changing set of users, devices, applications, and services. It needs to seamlessly and securely onboard this diverse set of devices and deliver the desired user and application experience.
It also needs to ensure fast and secure access to and between workloads wherever they reside. And for the network to function optimally, all this needs to be achieved end to end between users, devices, apps, and services across each network domain—campus, branch, remote/home, WAN, service provider, mobile, data center, hybrid cloud, and multicloud.
Which means that organizations need a new, integrated architecture for each network domain, one that is customized to meet the specific needs of that domain and that provides a way to communicate and enforce consistent policy across all domains.
The majority of today’s networks are not yet ready to meet the demands of this emerging digital era. In our 2019 Global Networking Trends Survey, we found that while 39% of IT leaders believe that their networks are very well aligned to meet the demands of digital business, only 19% of network strategists believe likewise.14
Yet there is reason for optimism. John Apostolopoulos, Cisco’s CTO for enterprise networking, anticipates a relatively short transition from today’s largely rigid, manually operated infrastructures to more agile, software-driven architectures that can “continuously adapt to meet the changing demands of the applications and services the organization depends on."
What will this emerging network architecture look like? According to Ravi Chandrasekaran, SVP of engineering, enterprise networking at Cisco, “Networks will operate as a system with increasing levels of autonomy, taking into account their own state, the dynamic state of all the users and applications, and the vast array of possible options."
Key to achieving this more autonomous state will be AI, which will help IT teams quickly respond to changing network conditions, whether that means automatically changing traffic routes, requesting more bandwidth, requiring a policy change, or even refusing a new service request.
Over time, by taking advantage of systemwide intelligence and automation, the network will become completely transparent to the user. It will simply be there, delivering secure connectivity to the services they need at the required level everywhere, at all times.
While Apostolopoulos admits there is still a long way to go before networks will have all the intelligence and power they need to deliver on this promise, he believes that the technical advancements needed to bring together AI-enabled service assurance, controller-based automation, natural language processing, and significant improvements in network security are well underway.
We found that while 39% of IT leaders believe that their networks are very well aligned to meet the demands of digital business, only 19% of network strategists believe likewise.14
In 2025, a leading-edge enterprise network will be able to take a requirement communicated in natural language from any line of business and automatically translate that into a set of policies and automated actions that will ensure that the business need is continuously met across the network—all without impacting any other existing services. A network with these kinds of capabilities is what’s commonly known as an intent-based network.
Here’s what a hypothetical use case for an intent-based network might look like.
Overview: An organization wants to use wireless IoT optical sensors to support a new business innovation delivered through an AR application. Here’s how business need and intent would be translated into network action.
Where do organizations stand today in their journey toward a more advanced network that can meet the demands of the digital era? Cisco’s digital network readiness model provides a standard five-stage maturity model to help IT organizations assess their current level of network readiness and help them plan where they need to be in the future.
The model can be applied across multiple network readiness categories such as architecture, access, WAN, assurance, network security, etc.
Answer these seven quick questions to see how you stack up against the rest of the industry.See how your network compares >
“By 2025, leading-edge networking teams will have intent-based networks operating across domains—campus, branch, WAN, data center, cloud, service provider, and security. Their networks will be able to comprehend business and application requirements and translate them to network and security policies. Agility will be dramatically improved through the network’s intelligent automation, and networks will operate with a powerful feedback loop that provides continuous monitoring, assurance, and optimization. The intent-based network will ensure that business services are continuously delivered and protected across the network. These advances will lead to significant benefits for organizations and also for society at large.”
— John Apostolopoulos, CTO for enterprise networking, Cisco