With AI-based networking, UNH Wi-Fi gets smart

In fall of 2017, the University of New Hampshire’s residence halls cut the Ethernet cord.

“Since the students no longer had wired ports in their rooms, we had to make sure the wireless network was rock-solid,” said Brian Simpson, information technologist at UNH. “To do that, we had to be a little more proactive.”

Around the same time, Aruba, a Hewlett Packard Enterprise company, approached the university’s IT team with an invitation to try its AI-based networking analytics tool as a betacustomer. They quickly found the service offered the granular, real-time visibility they needed to support the plethora of devices today’s students rely on for both work and play.

“They’re bringing Echo Dots, Amazon Fire TVs, Chromecast, Rokus — it’s a wide variety of devices we have to manage and support on the network to make students feel at home,” Simpson said.

After removing residence halls’ wired ports, Simpson’s team discovered that the school’s PC gaming enthusiasts had relied heavily on Ethernet for low-latency connectivity. The Wi-Fi simply couldn’t keep up.

Using Aruba’s AI-based networking analytics, the UNH IT team identified the time-sensitive gaming traffic and created rules to prioritize it.

Since the students no longer had wired ports in their rooms, we had to make sure the wireless network was rock-solid.

Brian SimpsonInformation technologist, UNH

“Before, we had to rely on broad views of the network, but NetInsight is a much more granular tool,” Simpson said. “We can pull up a historical, pinpointed report on an individual client and match it up to a trouble ticket. It gives us a much better on-the-ground view.”

The cloud-based analytics service also reviews access point configurations and compares them with APs in similar environments across Aruba’s customer base using machine learning.

In a couple of instances, Simpson said NetInsightflagged APs with overlapping coverage, causing clients to pingpong between them. The UNH team adjusted their wireless designs accordingly.

“It’s made a huge difference as far as coverage and capacity and the students being able to connect seamlessly,” Simpson said.

He added that the AI-based networking tool allows network managers to track building-by-building usage trends, informing long-term planning.

“It’s that sort of intelligence we were really looking for, so we can be proactive and get ahead of problems,” he said. “It’s been one of the most helpful tools we have at our disposal.”

[“source=searchnetworking.techtarget”]