Close

HPE Aruba Networking Blogs

HPE Aruba Networking Central: An AI networking perspective

By Frank Jas, Director of Product Management, HPE Aruba Networking

When the chance to be a part of the team responsible for reinventing HPE Aruba Networking Central popped up, I jumped at the opportunity. In addition to the look and feel, and a variety of other aspects, my Product Management team and I set out to speak to customers about how to leverage AI and automation.

Not surprisingly, the ideas we heard were great, but all were related to identifying problems — after the fact. While it is valuable to understand why access points, switches, gateways, and endpoints are having problems, there’s so much more. We wanted customers to understand that AI networking can be used outside of just troubleshooting tasks.

It’s my belief that a network management solution today should include optimization and endpoint security related insights, as well as troubleshooting assistance. By thinking more broadly and leveraging what we can learn from the network, and then combining it with endpoint telemetry, customers can gain from our management dashboard and AIOps capabilities without having to use and maintain multiple solutions.

Because of this, we capture very detailed telemetry from our infrastructure and all connecting endpoints. One of the strengths of our access points, switches, and gateways is the detailed telemetry each generates with respect to network traffic, client behavior, DPI, and application classification. By using our infrastructure with AI-powered HPE Aruba Networking Central, there is no need for separate collectors or dashboards.

In fact, a built-in AI networking feature called Client Insights is a prime example of how we’re utilizing the concept. Let’s say a new endpoint gets connected by someone in a remote location without IT’s knowledge – which happens quite often today. Using the telemetry gained from the network and endpoints with AI designed to aid in the profiling of each connecting endpoint, Central can tell that the endpoint is a surveillance camera, the brand and model, and if wired or wireless.

If it’s a known and approved entity, it’s easy to take advantage of defined roles, group tags and more. The best part is what happens next:

  • Automatic enforcement of access privileges
  • If a wired endpoint, the ability to dynamically assign the camera to a segment on an HPE Aruba Networking switch – no need to statically set up “camera” specific ports on every switch
  • Visibility into the traffic streams generated by the camera for security and bandwidth insights
  • And, because it’s a wired camera, you can receive a notification from Central if the camera exhibits abnormal behavior, such as an ability to roam (which may be a sign that the camera has been compromised)

If the fingerprint is not recognized and its role is “unknown”, the camera can be allowed to connect, or not. Your choice.

The last thing I’d like to leave you with is that when choosing HPE Aruba Networking Central, you get AI built-in – no added costs or setup. And because of the wealth of telemetry within our data lake and the constant re-training of our models, the network, security, and application-level insights that Central delivers offers unique insights that no other network management solution provides today.

For more insight:

What is AIOps?

What is AI-powered Device profiling?