adimas - Fotolia
The university, with more than 27,000 students, found that half as many HP Aruba APs were needed to cover the same area. The cost savings made the choice an easy one.
"Even with Cisco coming in cheaper per access point at double the count, why would I go there?" said Mike Carlin, CIO and vice chancellor for IT at UNCC.
HP provided UNCC as a customer reference for its Aruba wireless gear. As a first-time buyer, UNCC got a one-time discount for each AP. The university plans to buy 1,200 Aruba 320 series APs at $850 apiece. "They made it clear that's a one-time deal," Carlin said.
Cisco did not respond to a request for comment about the UNCC test. However, experts pointed out that the experience of a single university does not mean Cisco APs will come up short every time.
"Wi-Fi performance is hard to generalize, because the physical environment and client mix play very significant roles in real-world results," said Lisa Phifer, owner of consultancy Core Competence Inc., based in Santa Fe, N.M., which specializes in emerging Internet technologies.
While not the last word, the UNCC access point test is worth noting for companies considering a similar comparison. In June 2014, consulting firm Novarum tested the same two APs used in the university trial and found the Aruba gear performed much better than Cisco's product. Novarum compared the APs -- HP's Aruba AP-225 and Cisco's Aironet 3702i -- in real-world scenarios involving high-density throughput, client roaming and media streaming.
UNCC tests Cisco vs. HP Aruba
UNCC conducted its test last summer, with the vendors providing 28 APs each to cover two buildings -- one with offices and classrooms, the other one housing the university's food courts, Carlin said. The vendors chose their best 802.11ac Wave 1 APs for the test. Comparable Wave 2 products were not yet available.
"We found the Aruba APs had better range than the Cisco APs, which, in turn, gave us better performance," said Jesse Beauman, assistant vice chancellor for enterprise infrastructure at UNCC. "The Cisco APs also drained the mobile device batteries much faster for our users than the Aruba APs did."
The university then had the vendors provide plans detailing how they would use the APs to cover seven buildings based on a 5 GHz signal strength greater than minus 65 dBm (decibels relative to one milliwatt). The buildings included classrooms, offices, the student union and the library.
For capacity, the school asked that the vendors plan for five devices per person. UNCC also wanted the design to include the controller and management hardware, as well as the licensing for 2,500 APs. The controllers had to be redundant and support a minimum of 32,000 concurrent tunneled clients.
"This gave us a good sense as to what our future costs would look like, as well as insight into their design practices," Carlin said. "It was at this point we noticed that Cisco was consistently recommending twice the number of APs that Aruba was."
Cisco claims the Aironet 3702i performs much better than the Aruba AP-225. In a report posted on Cisco's website, Miercom, an independent product test center, found that the Aironet was six times faster than the Aruba AP.
UNCC runs its WLAN on Cisco Catalyst 3850 and 4506 switches, which have 1 GbE connections to the APs and 10 GbE uplinks to the school's data center.
The university plans to have half the campus on Aruba 320 series APs by early next year. Until funding is available, the rest of the campus will remain on the current 802.11n APs from Meru Networks, which security company Fortinet acquired this year.
Preparing your network for Wave 2 Wi-Fi
What Wi-Fi Aware means to the enterprise
Why the enterprise should focus on Wi-Fi capacity, not throughput
Check out HP Avaya's wireless access point line