Enterprise networking is seeing rapid and massive innovation: software-defined networking (SDN), artificial intelligence, the rise of the Internet of Things, and autonomous network actions are about to have dramatic effects on enterprise networking, from the data center to the network edge. These are the four network trends that every company should be aware of.

AI assistance

More and more network equipment vendors enhance their offerings by employing artificial intelligence (AI), mostly in the form of assisted machine learning (ML), to improve network management and security. ML algorithms continuously monitor the network traffic, looking for suspicious deviations from established traffic patterns. This allows IT and networking teams to be warned about early indicators of, for example, network performance degradation. The same kind of ML- or AI-assisted monitoring is used to scan the network for possible breaches or attackers. Threat actors often enter a network environment by getting hold of real end-user credentials and then abusing them for their own purposes. This way, they create deviations from the user’s previous network behavior — which ML can detect. Currently, however, a common problem in this scenario is that the detection mechanisms can only be as good as the data patterns the ML/AI system is fed.

Increasing autonomy

In the long term, networking will evolve to include better and, most of all, unassisted AI technology. This will give a boost to networking in several ways. First, it will allow the network to truly learn by itself, without the need for pre-established patterns fed by human hand. Second, it will enable the network to learn the reactions usually following a certain deviation. The network will, for example, recognize that privilege escalation (the abuse and extension of a legitimate user’s privileges by an attacker) will usually trigger certain reactions: the security team, once informed of the incident, will isolate the suspicious user and prepare that user’s network traffic history for forensic inspection. Once the network knows that an activity is suspicious, it will learn when to employ these steps autonomously — if the IT team allows it to proceed this way.

IoT enablement

The addition of IoT (internet of things) and IIoT (industrial IoT) technology to the network will create a whole array of new tasks, job profiles, and responsibilities. IT experts will have to deal with a wide selection of new endpoints. These include IIoT-enabled industrial machinery as well as smart factory and smart office technologies. The network will also have to interact with autonomous cars, along with various other kinds of smart and/or autonomous vehicles, as well as all kinds of smart devices will be added to the corporate network in the future. Combined with the AI capabilities an AI-enabled IoT network will, for example, allow farmers to send drones onto fields for applying pesticides in a highly targeted way, with multiple drones autonomously coordinating the work amongst themselves.

Software-defined everything

The increasingly dynamic, diverse and real-time-oriented networks of the future will be software-defined — just like some data center networks of forward-thinking enterprises and cloud service providers already are today. ‘Software-defined’ means that functionality is decoupled from the hardware, making the hardware ever more interchangeable. Just like in a virtual server environment, all important decisions are made in a software layer abstracted from hardware-specific code. This approach will be applied all over the future enterprise. It will be the task of the IT organization to define via settings, rules, and policies which kinds of software-defined network behavior they want to establish.

Taken together, the next generation network will pave the way for automating tasks that previously required time-consuming manual intervention. It will also enable new workflows that hadn’t been possible at all. This way, the network will speed up a huge number of existing workflows and enable whole new business processes – from the data center to the network edge and on to the cloud.

For example, a software-defined wide area network (SD-WAN) will allow a dynamic, policy-defined distribution of resources across the on-premises network and multiple cloud environments. It will support swift and easy relocation of resources across the multi-cloud environment, depending on current needs or shifts in the application workloads. SD-WAN functionality, by the way, is already available today. For, as science fiction author William Gibson once famously said: “The future is already here — it’s just not evenly distributed.”