When we first installed wireless at the University of St. Thomas (Saint Paul, Minnesota)in 2003, it was viewed as a patch to a very practical problem: offering our relatively few laptop-owning students a way to connect to the network. Today, however, all students arrive on campus expecting a fluid WiFi experience, from the dorms and the cafeterias to the classrooms and the quads. They expect WiFi to flow in the same way they expect lights to turn on when they flip a switch. Finding a way to fulfill this expectation has been an interesting journey.
For years, the “wireless sucks!”refrain was commonly heard on our campus, yet it was difficult to quantify and pinpoint our problem areas. Evidence of service deficiencies was largely anecdotal, and when we did enhance service, it was hard to quantify the impact. Heat maps showed us only the coverage; they didn’t help us understand the student experience. So, we chased down wireless ghosts, which did little to fix the problems students were having. We needed a way to find the real problems and to concretely demonstrate how improving wireless enhanced the student experience.
"Our job is no longer focused solely on monitoring the network to see if the hardware is up or down; we also focus both on verifying that the service quality meets our students’ expectations and on continually improving the user experience for all areas of our campus"
Overview: The Challenge
We introduced wireless at St. Thomas more than fifteen years go as a novelty item for students fortunate enough to own laptops so they could use the internet in classrooms lacking fully wired desks and workspaces. In those days, wireless was widely viewed as too unstable and difficult to centrally manage, while wires were better understood and more reliable. As a result, we focused our wireless squarely on academic areas—mainly classrooms—because it was impossible to imagine a business need for this novel technology in actual working environments.
How wrong we were.
Today, nearly 10,000 wireless devices connect to more than 1,300 access points at St. Thomas each day, and the environment can’t grow fast enough. As students bring more and more wireless devicesto campus, the appetite for bandwidth and reliable connectivityis all but insatiable. From laptops to smart phones, tablets, gaming devices, digital assistants, media players, and smart TVs,the number of devices using wireless on campus has exploded, and students expect everything to work flawlessly.
Enter the Network Engineers
With little if any control over how all of these devices behave (or misbehave), industry has left it to network engineers to make everything work together using a magical and little understood service. This service is pervasive in most of our students’ homes before they arrive on our campus—their “home away from home”—and they expect it to be everywhere at all times of the day or night, just like at their parents’ house. Indeed, one of our colleagues quipped that, for many students, wireless is more important than running water.
So, how do you create such a wireless experience in an effective and efficient way? This was our challenge. And it seemed nearly insurmountable until a member of our networking team discovered a new tool all but tailored to fill the gaps in our user experiences.
This product not only monitored the devices that we controlled and that were connected directly to our network, it also analyzedour wireless controller logs and network traffic to help us gain visibility into something previously off-limits: the user and device experience. And, while other products out there did log aggregation and reporting, none of them made the process as easy and non-intrusive as the tool we chose: Nyansa’s Voyance Campus, which let us follow the hops on our network right down to the devices themselves.
Being able to tell how well (or poorly) a device was connecting to our wireless network—without having a student march the device across campus or our having to conduct a site-visit—was immediately helpful. What was alarming was the truth about our wireless infrastructure, which the analytics also exposed. Plain and simple, the coverage redesign implemented in 2007 was inadequate to deliver what users expected in 2018.
Previously, we would hear about “bad WiFi” through various means, but those reports rarely gave us any useful data with which to make informed decisions. Now, we can see the poor quality and channel interference directly. By analyzing device traffic, we can quickly determine problems with user passwords or issues with a particular device’s operating system version or settings, and easily visualize where actual wireless network problems exist…and where they don’t. This visibility let us make immediate improvements. After only a few days of initial analysis, we were able to focus our upgrades in the worst wireless locations on campus.
We quickly put a plan together to tackle the residence hall with the worst results. Analysis after that building was upgraded showed that our work was paying off. Not only did the support calls all but disappear from the upgraded building, but we were also able to see
o more devices connecting at 5 Ghz,
o less bouncing between access points, and
o a better overall balance of radio power adjustments throughout the first residence hall we upgraded.
Similar work on another residence hall yielded the same results. Our work was giving users a better experience right away, and the result was an immediate drop in trouble tickets to our Help Desk.
Visible Evidence of Dramatic Change
So, after pulling together additional information, we made our case: The network itself wasn’t the issue, per se; the problem was in how wireless network usage had changed over the past 15+ years.Dozens of smaller mobile devices in tight quarters with bandwidth-hungry gaming systems and computers were appearing in places that may have had only a few 2.4 Ghz-only laptops years ago, yet our tired wireless network trudged on unaltered. And, while we still needed to extend it into areas we never dreamed of, coverage was no longer our core problem; quality of service and density mattered more than ever.
Because the results of Voyance Campus analytics are visible to even the most engineering-challenged observer, additional funding is on its way.We plan to have all the wireless brought up to modern standards in the residence halls by the start of the 2019–2020 school year. The remaining academic and administrative buildings will follow—and will require additional capital investment. As we complete each phase, we will be able to concretely see where we have succeeded and where we must do more.
In today’s world, “WiFi Available” signs are no longer necessary; as such service is now an expectation rather than a luxury or a nice-to-have. As a result, our job is no longer focused solely on monitoring the network to see if the hardware is up or down; we also focus both on verifying that the service quality meets our students’ expectations and on continually improving the user experiencefor all areas of our campus.