Networking & Wireless

How Do We Make School Network Infrastructure Affordable?

As districts adopt 1-to-1 or BYOD, they need to adapt new ways of financing and maintaining their networks.

• By Marie Bjerede, Keith R. Krueger
• 07/09/15

This is the second article in a three-part series on network infrastructure. You can read part 1 of the series here.

"How do we pay for it?"

This is often the first question that trailblazing districts implementing 1-to-1 programs hear from those hoping to follow in their footsteps. It's typically followed by "What devices should we buy?"

According to districts that have already implemented mobile learning, these are the wrong questions. They come from an outmoded framework that doesn't reflect just how dramatically and thoroughly everything changes once a district makes the decision to pursue a digital transformation. Instead of questions about costs and devices, districts should be asking themselves, "Why do we think implementing 1-to-1 is important?" or "What will teaching and learning with technology look like here?"

When a district begins with these core questions, the answers to other questions will naturally ensue. Leaders will be free to choose devices based on the learning tasks for which students will be using them, then holistically restructure budgets as new ways of working completely change the tasks and priorities of the IT and instructional organizations within the district. Many of the jobs that belonged to IT will go away as it shifts to a service organization, eliminating old job descriptions while creating new ones. Curricula and instructional approaches will change, and money will shift toward devices and digital tools rather than textbooks. Many districts are using these wholesale changes in the day-to-day work of the district to implement 1-to-1 without budget increases.

The Impact of 1-to-1 on Networks
Infrastructure investments, though they entail much higher upfront capital costs, can be addressed using a similar, system-wide, holistic approach. When a district embraces 1-to-1, and particularly BYOD, the demands on the network go non-linear. Leading districts find that they experience exponential growth in their need for network capacity. This growth varies from district to district and depends on how technology is actually used and implemented, but it is common to see 60 percent growth in demand on network capacity each year — meaning that demand on the education network doubles every 18 months.

Users' expectations of network reliability also increase dramatically as districts move to 1-to-1 computing strategies. Where it used to be merely inconvenient for the network to go down (when it was used primarily for operational needs such as maintaining student records), once teachers and students begin to depend on the network for their daily work, it becomes mission-critical. District leaders report that their administrators, teachers and students expect essentially zero downtime, yet CoSN's E-Rate and Infrastructure Survey indicates that more than one-third of districts reported more than three days of downtime annually. Shockingly, 5 percent of school districts reported that their education networks are down 30 or more days a year, more than one day of instructional time each week.

Districts also need substantially increased ability to support more devices. When many school networks were established, they were intended to support at most one device per adult. Once districts move to 1-to-1, the number of devices may increase by more than 20 times. Where BYOD is allowed or encouraged, some districts are seeing three or more devices per student. This means that the new networks need to support almost two orders of magnitude more devices than their old networks did.

Nearly all districts will have to accommodate 10 to 100 times the number of devices when moving from computer labs to 1-to-1, with their unique circumstances driving where they fall on the spectrum. This change has important implications for addressing scalability of network appliances and security models.

Increasing Demand Means Increasing Cost and Complexity
Dramatically increasing requirements creates a critical dilemma for districts in planning and budgeting for infrastructure purchases, installation and maintenance in a number of ways.

• Network capacity is expensive. According to the CoSN survey, one-third of districts are paying more than $50 per month per Mbps (although a fortunate 27 percent of districts pay less than 10 percent of that cost.) If a school is trying to achieve the Federal Communications Commission's long-term recommendations of 1 Mbps per student, this means that a district of 10,000 students may have a hefty monthly bill of $500,000. Even at the 2014 targets of 0.1 Mbps per student, the cost could be $50,000 per month.

• Network appliances are expensive. If districts expect to see a doubling of capacity demand every 18 months, the 1 gigabit appliances they currently use for for filtering, switching and security will likely become obsolete. If so, they will need to update to 10 gigabit versions of the same appliances well before their give-year technology plan expires and new bonds or other replacement funds become available.

• Networks may be unreliable. Many networks evolved under the assumption that most computer access would be to servers located in the school or at the district headquarters. These networks have the traditional hub-and-spoke architecture, with a single line going from each school to the district, and the district then connecting to the commodity Internet. In this model, each connection is a single point of failure where an outage will affect a whole school, or even the whole district, until it is repaired.

• Network design is complex. Designing highly reliable, high-capacity networks calls for professional expertise that many districts can't afford to maintain in-house.

• Network maintenance is complex. Managing networks for performance, reliability and capacity requires specialized expertise that, once again, may be difficult for many districts to afford or even find in small or rural communities.

Fortunately, many districts can address these challenges by mitigating market factors and mitigating approaches.

Funding and Logistical Solutions
First, the cost of network hardware and network capacity are on downward sloping curves. It is possible for districts to plan carefully for the capacity growth that drives hardware infrastructure investments, making judgments about how long they can afford to wait to upgrade appliances and balancing their needs against how costs decrease over time. As for network capacity, many districts find that though their costs stay constant, the amount of capacity they are able to provide over time increases to keep pace with demand.

Second, E-rate modernization has made it easier for districts to fund infrastructure to close the WiFi gap. E-rate has added $1.5 billion funding per year for school broadband connectivity, with $2 billion available this year. E-rate now also supports long-term planning for infrastructure by making it possible for districts to receive funding over a five-year period rather than spending the full amount within one fiscal year.

Third, states and large districts are finding that they can dramatically reduce costs through the purchasing power they wield when they aggregate demand across districts or schools. This aggregation is effective not only for purchasing hardware and capacity, but also for getting the technical human resources required for professional network architecture development, network design and implementation and network management and maintenance. To make this approach easier, the FCC is encouraging consortia and bulk purchasing by removing constraints that previously prevented districts who were seeking E-rate funding from participating with other public sector entities such as rural healthcare providers, universities, counties and municipalities.

At the state level, the intrinsic scale of multiple school districts makes it possible to negotiate for the best pricing on hardware and bandwidth. The same is true for large districts serving tens of thousands of students. Equally important, these large organizations are able to support high-quality network engineers to design and maintain their networks. For smaller and rural districts that are not fortunate enough to have access to state-supported services, other aggregation strategies become critical.

Aside from purchasing consortia, districts can cooperate to get network architecture and design support. This work is critical during times of build-out and investment, but there are significant logistical and practical challenges to realizing those benefits.

As an alternative, many districts are finding that it is more cost-effective to contract for their networks as a service, using third parties that provide network architecture and infrastructure build-out as well as day-to-day network management and active maintenance and scaling. In the cases where operational expenditures are easier to budget and plan for than capital expenditures, the whole network, including devices, may be obtained as a service or lease. These districts generally enjoy automatic device refreshes as well as upgraded bandwidth and appliances when needed.

Finally, districts are beginning to take advantage of market aggregation by building a carrier-neutral data center where colocation, telecommunications, Internet and cloud services are available at much higher levels of service and much reduced cost.

By taking advantage of the collective power of consortia, collaboration with other districts, market aggregation and/or third party services, even small districts gain the power to ask — and cost-effectively answer — the right questions.