Over the past 5 years or so we’ve definitely seen a fair amount of change when it comes to enterprise-grade Wi-Fi access points.
Today, the latest and greatest is 802.11ac Wave 2, however, before you go out and replace every AP you currently have with the new wave 2 technology let’s take a closer look at the latest standard to understand exactly which access point is right for your environment.
As part of our new wireless network design basics series, we sat down with our very own Michael McNamee (Mobility Practice Manager and Senior Network Engineer at SecurEdge Networks) to have an in-depth discussion about 802.11ac access points and what IT managers should consider before their next WLAN update or refresh.
SecurEdge Networks: How many users can an 802.11ac access point really handle?
Michael McNamee: It Depends. This is a tricky question because there are variables in this equation that must be taken into account.
First, because Wi-Fi is a shared medium all connected devices are given a slice of time which we refer to as air time. The more devices connected to an access point the less air time or the smaller slice of time each user is able to access.
As you add more devices to an AP the potential for slower performance increases because everyone has a smaller slice of time. The end user experience results in slower response times to page refreshes or downloads of files and streaming media may stutter or buffer.
Second, the types of devices that the end users are connecting with also have an impact on this capacity. Each device type; laptop, tablet, phone has a profile of what data rates it supports as well as how many transmit TX and receive RX streams it can use. These capabilities affect Wi-Fi capacity in the way that they correlate directly to air time.
For instance a phone or tablet (1 TX/RX spatial stream) can demand up to 3x the airtime that a business laptop (3 TX/RX spatial streams) would demand. The net result of this is if I have to accommodate all laptops on an AP then I can have many more of those than if I had to host all tablets. In essence the tablets are slower and consume more of the slices of time on Wi-Fi.
Third, the applications being used also have to be accounted for in factoring for AP capacity. If the application is simple text to an inventory application or an EMR application the bandwidth requirement is low. However, if the application is an HD video or large medical scan then the bandwidth demands are going to be much higher and will result in fewer devices that can be accommodated at any one time on the AP.
Fourth, we have to look at the RF spectrum and which frequency we are using. With Wi-Fi we have two frequency bands to use; 2.4GHz and 5GHz. In each band we have a limited number of channels or frequency slots that we can use. In the 2.4GHz band we have 3 non-overlapping channels that we can use. In the 5GHz band we have up to 22 non-overlapping channels available for us to use.
Due to the few channels in 2.4GHz we cannot take advantage of channel bonding to provide higher data rates for increased bandwidth. In the 5GHz band however we can bond multiple 20MHz channels together to get much higher data rates and thus increase the bandwidth provided to client devices. This means if my AP can support 5GHz and there are 5GHz connected clients I can potentially accommodate many of them because I have more bandwidth to provide.
My response of “It depends” posits that the dependency is based on the variables posed above. We need to know how many potential users there might be, what devices they are connecting with, what applications they are intending to use and what Wi-Fi bands are going to be utilized.
Once we know all these variables we can run some calculations using some great tools like Revolution Wi-Fi’s Capacity Planner to better understand the demand that is going to be presented and what we need to provide to accommodate them.
802.11ac however presents us with more bandwidth than we have ever had before and generally not only has the technology gotten better but so has the hardware.
With the improved hardware and technology we typically see 20-40% better capacity gains on APs than we did with 802.11n. Again, take this observation with a grain of salt and realize our experience is with Enterprise grade equipment and not the consumer grade APs you can buy online or at your local office supply store.
SecurEdge Networks: 802.11ac Wave 1 vs Wave 2 which is more practical?
Michael McNamee: Right now as of this conversation I would say Wave 1 is the more practical of the two. Wave 1 has been out for a full 2+ years, the technology is mature and there is a great selection of client devices that support 802.11ac.
The devices that support 802.11ac Wave 1 can take advantage right away of the better performance it offers. Also, the Wave 2 APs that are available right now are the high end, flagship APs and are the more expensive models which make them a bit impractical for most situations.
SecurEdge Networks: Are there any scenarios where wave 2 would be a better fit?
Michael McNamee: The differences with Wave 2 and what it offers is really only a few key features. First we get the Multiple User - Multiple In Multiple Out (MU-MIMO) which solves the one device at a time constraint that current 802.11 Wi-Fi technology harnesses us with. Wave 2 MU-MIMO capable APs will be able to communicate with multiple clients at a time.
This promises to triple or quadruple the speeds at which Wi-Fi operates. However, in order for this to work, the client devices must also support MU-MIMO and at the moment there aren’t many of those devices out there.
The other feature is the ability to bond up to 8 channels in the 5GHz radio band to give us better bandwidth. Realistically this would never occur in most Enterprise deployments where more than 2 APs would be deployed.
Ideally Wave 2 would be a good fit for a home environment where there would not typically be more than 2 APs so the use of 80-160MHz channels and the reduction of available channels is not a challenge. The other environment would be one in which all the devices deployed are MU-MIMO capable and needing the additional bandwidth. Wave 2 in my opinion is going to only apply in very niche situations until MU-MIMO capable devices become the norm.
SecurEdge Networks: Which type of environments would you recommend deploying 802.11 ac?
Michael McNamee: ALL of them. Since 802.11ac right now is the latest technology and is the same cost as 802.11n I would advise going with 802.11ac. Also, we’ve seen in every deployment that we’ve done of 802.11ac APs that it has improved wireless performance all around.
Even in instances where the client devices are still 802.11n we have witnessed better performance when the 802.11n device was connected to an 802.11ac AP than when the 802.11n device was connected to an 802.11n AP. This surprised us at first since we weren’t expecting it but we quickly began showing customers what the impact was to performance when moving to 802.11ac.
SecurEdge Networks: What’s the best approach to deploying 802.11ac access points? (Focusing on our design process)
Michael McNamee: If you are making the jump from legacy wireless 802.11a, b, g to 802.11ac I would highly recommend a revisit of your wireless design by a WLAN design professional.
When we migrated from the legacy 802.11a, b, g technology of diversity antennas to 802.11n MIMO we had to change our design strategies on where to place APs. The migration from 802.11a, b, g taught us a lot in the very beginning and after adjusting our design approach we saw marked improvements in customers Wi-Fi networks.
If your current Wi-Fi network is 802.11n and you made those design changes already then it is pretty much a remove and replace of the old 802.11n equipment. However, I am always going to advise that you consult an experienced WLAN design professional when making any changes in WLAN technology.