As you get ready to procure a wireless system, there are a few key decisions necessary to ensure that you get a system that meets your district's wireless needs. These are the decisions that ultimately drive the performance and cost of potential solutions. If you don't have time to consider every feature included in this guide, focus your time on this shortlist of key decisions.
Most modern APs have at least two radios that transmit and receive client traffic. One of the radios operates in the 2.4 GHz band, and the other in the 5 GHz band. An AP needs to have a radio operating in each of these bands to be considered a "dual band" wireless AP.
The 2.4 GHz band is significantly more crowded and prone to interference than the 5 GHz band. Whenever possible, clients (laptops, tablets, etc.) should be pushed to use the 5 GHz band for better overall system performance. Though single band (2.4 GHz only) APs are still available on the market and are less expensive than dual band, the cost savings is never worth the performance tradeoff.
However, signals of different frequencies also have different propagation characteristics. Signals with lower frequencies (2.4 GHz) are able to travel further distances and propagate through objects like walls and doors better than higher frequencies (5 GHz). Depending on the material type, whether it is drywall, cement, rebar, wood, or another material, 2.4 GHz signal may propagate through the wall but the 5 GHz signal will not. The denser the material, the less signal will leak through no matter the frequency.
EducationSuperHighway strongly encourages you to buy dual band APs rather than single band for optimal performance and to accommodate a large number and variety of clients.
Keep in mind that because of the higher frequency, the 5 GHz signal does not propagate through barriers (walls, doors, etc.) as well as the 2.4 GHz signal. If your wireless network today was designed for coverage by 2.4 GHz radios, you will need to purchase more dual band APs than are currently in your network for proper coverage by the 5 GHz radios as well.
With the recent release of 802.11ac Wave 1 and Wave 2 APs, the market is at an inflection point. Customers must decide whether to invest in the new technology now or for future generations of APs to be released. This is a difficult decision and each district will differ depending on the existing network and its intended future use. The district must consider the existing density, bandwidth needs, and hardware of the clients now and over the next three to five years to determine whether the additional throughput from an 802.11ac Wave 2 AP is necessary.
Client device type plays a role in the type of network you need. If all of your devices only support 802.11n and you do not plan on introducing 802.11ac clients within the lifetime of the new APs (3-5 years), an 802.11ac network should not be a priority. If you do plan on introducing 802.11ac clients within the next three to five years, they will be able to take advantage of the enhanced performance features of 802.11ac. For clients to take advantage of the new features in both Wave 1 and Wave 2, their hardware must be compatible. Many vendors continue to flood the market with 802.11n capable devices, rather than 802.11ac capable ones which has kept consumer costs down. Newer devices that support 802.11ac Wave 1 have recently become affordable for school districts to purchase. Wave 2 capable devices are not yet common and can be expensive. Timing of device refreshes and network upgrades alignments should be considered when choosing between Wave 1 and Wave 2. You do not want to rush to adopt Wave 2 and incur extra expense if your client devices will likely not have the ability to take advantage of the new features for a few years.
The cost of most technology drops with the introduction of new iterations or improved versions. In most cases Wave 1 APs are less expensive than newly introduced Wave 2 APs. It could take a few years of adoption before the price of Wave 2 drops significantly. With this in mind and also factoring in your client refresh plans for the next few years, investing in Wave 1 technology may be a better choice based on costs and client refresh cycles. With this in mind, it may be smart to consider 802.11ac Wave 2, not for this upgrade cycle, but for the next.
Advancing your network from 802.11n to 802.11ac is highly advisable. 802.11ac Wave 1 AP prices have become extremely competitive in comparison to both 802.11n and 802.11ac Wave 2 APs. If your district’s educational technology goals are geared toward a 1:1 or media rich environment within the next four years and you will have 802.11ac devices on the network, EducationSuperHighway recommends upgrading to 802.11ac Wave 1. Otherwise, if you have all 802.11n clients and an existing 802.11n network, you do not have aggressive technology goals over the next 3-5 years and you are on a tight budget, 802.11n (three spatial streams) will likely suit your needs and you may wait to upgrade. Also consider the advertised speed of each standard and what port speeds your wired switches ultimately support as there can be a cost impact for both the AP and switching gear needed. Bottom line, buy what you can afford.
Until recently most APs had a single Gigabit Ethernet (GbE) port to backhaul data. With the release of 802.11ac, many manufacturers are installing a second GbE port in their APs because of the advertised 1.3 Gbps speeds.
Many manufacturers are adding APs to their product line with a second GbE port and charging a premium for those "higher performing" models. Don’t be fooled by this.
The theoretical throughput of a typical 802.11ac AP is 1.3 Gbps on the 5 GHz radio + 217 Mbps on the 2.4 GHz radio, so the common misconception is that a second GbE port is required. However, this logic is based on theoretical data rates of the two radios. The real throughput (which is typically ~60% of theoretical) of the 802.11ac Wave 1 AP will not exceed 1 Gbps.
Do not choose a more expensive AP if the only added functionality is a second GbE port. These are the only situations where a second Ethernet port is needed:
Until the release of 802.11ac, APs operated at full performance on 802.3af power (15.4w), commonly known as PoE. Many 802.11ac APs require more than 15w to operate at full performance and require 802.3at (25.5w), commonly known as PoE+.
If most 802.11ac APs are installed in a network with PoE switches, the new APs will not receive enough power to run at full performance. The APs will function, but they will start running in low power mode so you will not see expected performance from them.
Ideally, if you are purchasing 802.11ac APs that require PoE+, you should consider upgrading your switching infrastructure to PoE+ at the same time.
If you are upgrading your wireless network to 802.11ac but plan on leaving your PoE switching infrastructure in place, you should talk with the manufacturer to see what performance implications it will have. There are some 802.11ac APs that run full performance on PoE, but making that a requirement will limit the APs that are eligible. Instead, you may want to purchase a few PoE+ switches and plug the 802.11ac APs located in high density areas (like the library or assessment rooms) into the higher power switches.
If your district currently has 1:1 learning or is planning to move to a 1:1 environment within the next 3-5 years, the question on your mind is probably "What type of wireless system do I need to support 1:1?" Luckily, most wireless solutions, if installed with the correct density can support a 1:1 environment.
There is a lot of debate about the correct number of APs per classroom. Many 1:1 districts have installed less than one AP per classroom in a saw tooth pattern and have achieved perfectly acceptable performance. Other districts swear that one AP per classroom is needed to handle their constantly growing bandwidth demands. Ultimately, it all comes down to a good site survey. Without the site survey, which takes all of the variables into consideration (building construction, number of user devices, type of user traffic, interference levels, 802.11n vs 802.11ac, and more), it is difficult to make a blanket statement about how many APs every district in the country needs.
Many districts have found that, with a tight budget, they have had to start with fewer APs to provide coverage instead of density (one AP per every two or three classrooms), and begin to increase density each year as they are able to purchase more APs. This strategy is fine as long as the vendor is aware of this strategy when performing the site survey such that they can suggest an appropriate strategy to add APs each year.
EducationSuperHighway has recommendations (see chart below) about the number of APs per classroom that a typical district may need for high-density Wi-Fi. These recommendations are meant to be used as planning and budgeting guidelines, but they are not intended to replace a good site survey and should not be used as a specification in your RFP otherwise you may end up buying more than you need. Your RFP should require that the vendor perform a site survey and base the design on their results.
Modern wireless networks all implement the concept of centralized control of the APs. The centralized control can be done via a physical controller or a cloud-based controller (for now, let’s say that controller-less and virtual controller systems fall in this category). Physical controller architectures have been around for years and have been proven to be very stable and secure. Cloud-based architectures are newer but have become very popular over the last few years among school districts because of their simplicity. Either are good options for districts of any size.
Physical controller architectures benefit from having somewhat more local control over the user traffic. The controller can process every packet of data, so some administrators consider this design to be more secure and better at providing visibility into the traffic. Also, the physical controller is situated within your district’s network, so you are not relinquishing control of uptime or usage data to the manufacturer hosting the controller.
Cloud-based solutions rely on the manufacturer who is hosting the controller to maintain a high level of uptime. While this was initially a concern for customers, the uptimes have been extraordinary so this argument has become less important over time. The main benefit of cloud-based solutions is their simplicity to install, configure, and maintain. They also tend to have more straightforward purchasing and licensing models.
If you have an existing physical controller infrastructure that you like or if you have controllers that are not yet at their end of life, you may want to take advantage of your existing investment and continue with a physical controller architecture. Or, if you like the added control and security of a physical controller you should consider a physical controller architecture.
If you are not really sure where to start or are looking for a quick and easy solution while still having a straightforward management experience, a cloud-based architecture may be the best approach for you. However, due to high ongoing licensing costs, these solutions may not be an option for districts on an extremely tight budget.
Ideally, you should talk to your vendor to discuss the architectures and their associated price tags might be best for you.
Depending on what you decide, you will still need to make some deeper architecture decisions. Continue to either Which physical controller based architecture is right for you? or Should you consider a controller-less or cloud based controller architecture?
In this architecture, the controller(s) resides at the district office (or any central location) and all of the remote school APs are controlled across the wide area network (WAN). If a district has multiple small schools, a central district office, and a reliable high-speed WAN with low latency, a centralized controller model may be a cost effective design.
In this architecture, controllers reside at each school (or at large schools) rather than centrally at the district office. If a district has large schools or an unreliable WAN this is a smart choice.
Recently "converged access" is becoming more prevalent. This means that the switch acts as a local controller for directly attached APs.
Consider your traffic flows, stability and capacity of WAN, content caching, budget, management capability, and high priority features before choosing a controller architecture.
Every district has very different WAN infrastructure, requirements, and budget so your choice of architectures should take these into consideration. If you have a highly reliable and low latency WAN, you can consider a centralized controller solution because it is easier to manage and less expensive. If you do not have a reliable WAN, you should consider a distributed controller but be aware that this will increase the cost of the network.
Ultimately, it is important to understand that every school district is different. We recommend that you seek input from your vendor before making a decision.
Autonomous APs (frequently known as heavy or fat APs) maintain their own individual configuration and do not receive configuration changes from a controller.
Sometimes called the distributed controller architecture, coordinated control functionality (equivalent to what a controller provides) is split up amongst all of the APs. This may sound similar to the standalone architecture, but controller-less systems are much more sophisticated such that the APs work together to create a dynamic network, much like in a controller environment. Though the controller functionality is a part of the AP, these solutions typically use management software to manage and monitor the APs and clients.
There are a limited number of controller-less systems on the market, but the ones that exist are very good and should not be ignored simply because they do not use a controller. Examples of controller-less systems are:
A cloud controller is a controller that is hosted by the manufacturer and your configuration/AP management are all done by logging into your district’s management interface using a web browser.
A virtual controller is controller software that can be installed on any system, either locally or cloud based. This controller does not necessarily have to be available 24/7 to manage the network.
Cloud based controllers are becoming more popular so many manufacturers are starting to release their own cloud product lines. It may take some time for these new products to stabilize, so be sure to check for customer references if you are considering one of the newer products.
Standalone AP architectures are outdated and do not provide nearly the same number of features, functionality, or performance than other types of architectures. No matter how small the wireless installation, it is worth the extra money for the advanced management features that come with the either a controller-less architecture or a cloud-based controller.
The differences between the controller-less and cloud-based controller systems are all "under the hood" so users and administrators will generally not be able to tell the difference between the two. Both of these designs are good options for districts of any size. They are straightforward to design, install, manage, and monitor. It is important to be aware that both may require a per AP license so be sure to ask about this when exploring these options.
Most 802.11ac Wave 1 APs on the market today support at least three SS so the two versus three SS decision only comes into play when purchasing 802.11n APs. Wave 2 APs support up to 4 spatial streams.
Choosing an AP with three versus four SS is typically a cost versus performance decision. The additional hardware and intelligence for the SS increases the price of the AP, but also increases the performance.
You should also consider the type of client devices that will be on the wireless network. Smartphones typically only support one SS, tablets support one or two SS, and laptops two or three SS.
If purchasing an 802.11ac Wave 1 network, three SS is the norm. Wave 2 APs can support four spatial streams. While there are many clients in use today that only support two SS, the cost versus performance tradeoff is usually not large enough to warrant a four SS network. However, if the majority of the devices are two SS devices, with newer devices supporting a third SS, prioritizing a fourth SS should not be as high a priority if you have a tight budget.
Many manufacturers are moving toward a unified access layer - meaning access switches and wireless access can be configured and managed via the same platform.
This collapsing of management platforms can save the district time and money both in the purchase of physical appliances or software and time training staff to use the management platform. However, the benefits of the unified access management are not so great that a district should consider replacing switching infrastructure simply to accomplish this unification.
If your district has an up-to-date switching infrastructure, keep it. If your district is upgrading both the switching and wireless infrastructure at the same time, unified access layer management is worth considering but should not weigh too heavily in your decision.