4 tips for optimizing wi-fi on campus
director of product marketing, Mist Systems
Want to make a college student laugh? Hand them an Ethernet cable.
For today’s students, it’s all about the wi-fi. They’re bringing an ever-growing assortment of internet-connected devices to school—phones, tablets, laptops, gaming consoles, smart speakers, fitness trackers etc.—and they expect fast, ubiquitous wireless access throughout campus.
The enormous challenge of meeting those expectations extends beyond students. Wi-fi also is essential to the faculty’s increasing use of online tools for enriching the learning experience, as well as realizing the highly automated smart campuses of the future that many universities envision.
But there’s a big problem: Campus WLANs weren’t set up to handle thousands of people accessing the wi-fi with multiple devices in different places all at once. Built on aging and obsolete architectures, these networks are too often prone to deliver the unreliable wireless that students hate as much as final exams and financial aid forms.
Out of this crucible, fortunately, new technologies and practices are emerging to turn wi-fi from a nagging headache for higher ed institutions to a quality-of-life enhancer, recruitment selling point and IT efficiency engine. This includes the use of artificial intelligence (AI) to watch over the network and pinpoint service issues much faster and on a far greater scale than is possible with traditional, labor-intensive WLANs.
With that in mind, here are four principles that every university should follow to ensure the best possible wireless experience:
1. Adopt AI for troubleshooting at scale in real time.
It’s incredible to think that, despite the heavy burdens on WLANs today, the network architectures in place on most campuses were designed more than a decade ago, before the popularization of the smart phone, cloud computing and social media and when Netflix meant mailed DVDs.
These old networks are dumb in that they provide little visibility into how users are experiencing the service and, when something goes wrong, administrators must manually comb through a hodgepodge of computer-generated logs to figure out the issue.
New AI-driven systems, however, are much more efficient at troubleshooting than humans could ever hope to be. Dartmouth College, for example, is using such an AI for IT system to collect vast amounts of information about what wired and wireless users are experiencing on any device, analyze that data in real time in the cloud, and instantly zero in on where exactly in the network a problem started and why. This is true AI in that the systems learn over time and can predict problems even before they occur.
The technology that Dartmouth is using even offers a Siri-like “virtual network assistant” that uses Natural Language Processing to announce, for example, “Your DHCP server didn’t respond” and then walk administrators through determining the reason.
As Dartmouth is discovering, AI can be a savior for overwhelmed university IT staff and help desks.
2. Design for capacity, not coverage.
Back in the dark ages, when little was asked of wi-fi beyond connecting a laptop to check email, universities tended to spread out access points (APs) across a defined area (typically distributing the fewest possible APs to provide sufficient coverage).
Today, users are going wireless on more devices with more applications and heavier workloads, yet they expect wi-fi to perform at the level they are accustomed to from a wired network. Thus, planning for capacity has become more important than coverage.
Universities’ credo should be “capacity everywhere”–ensuring the WLAN can provide optimal performance wherever students congregate, for whatever they’re doing–whether it’s online research or streaming Hulu. It’s crucial to think about exactly how and where users actually use the network.
That means weighing a variety of factors–including the number and types of end user devices, the breadth of wi-fi protocols and RF bands that the clients support (e.g., wi-fi 6 in the future, 802.11ac, 802.11n and even the ancient 802.11b on older devices that still haunt the college halls and labs), the bandwidth consumption of most popular applications, and the daily traffic patterns of students–and smartly deploying APs accordingly.
For example, it doesn’t make sense to put an AP in a hallway (which might have worked fine in a coverage model), when the needs of 200 students in a large lecture hall should dictate more APs inside the room. And remember that the same lecture hall could need to serve twice as many users when students waiting outside for the next class are accounted for. Planning for today’s WLAN demands requires this high a level of granularity.
And as the bandwidth needs of devices and capabilities of the APs continue to soar, you may even need to consider migrating away from a centralized data plane that is inherent in legacy wireless controller architectures.
3. Plan for the smart campus of the future.
Technology is changing the campus experience.
The Bluetooth Low Energy (BLE) capabilities in everyone’s smart phones can interact with wireless beacons to provide a range of new mobile services such as step-by-step directions for college tours or lost freshmen inside large academic buildings, automatic taking of attendance, and push notifications for a special promotion in the campus bookstore.
Meanwhile, universities are exploring Internet of Things applications such as smart door locks in residence halls and heating and cooling systems that can intelligently adjust themselves based on the number of people nearby to save energy and optimize comfort.
This new wave of location-based services will only increase the need for superior wireless. Higher ed institutions must take this into account when deciding their WLAN requirements moving forward.
4. Embrace APIs.
An AI bonus? The ability to extend the technology in new ways by leveraging its APIs.
For example, Dartmouth College is creating a self-service portal that allows students, faculty and staff to report problems. It integrates with the automated troubleshooting capabilities and messages users that the issue has been resolved. The college is also using an API to provision APs and upgrade firmware.
All universities would be wise to imagine the kinds of creative software-based solutions available in modern, AI-driven WLANs.
By following these four tips, higher ed institutions can meet the challenge of delivering wireless internet anywhere, anytime on campus and ensure a productive future.