Wi-Fi Design, 802.11ac, And Planning For What’s Next
We are talking about Wi-Fi design.... again.
This is not the first article about Wi-Fi design, and not the last one. Definitely not the last one. The reason being is, things change: new technology, new implementations, new use cases, evolving environments.
This is why I often feel amused when listening to, or reading the opinions of, industry leaders on various topics - Radio Resource Management (RRM), physical AP placement, capacity, authentication methods are but a few examples of that.
I often feel like I am watching a track and field 400m hurdle race where experts take a snapshot of a group of athletes going over hurdle n.3 and discussing the best and most efficient technique.
Don't get me wrong, these experts are not experts just in title, they know their stuff. But at the same time the race is already at hurdle n.6 and the order of the athletes has changed. The industry is not stopping, and the design is changing (to a degree, at least the RF still behaves the same so know your basics).
"Wave2" 802.11ac technology, higher client capacity requirements, incompatible technology entering the unlicensed band, and added emphasis on visibility, and control of the wireless network are the biggest design influencing points.
There is a big caveat when designing for above. Everything is NOT changing at the same time. Designing "green field" deployments is fun, but quite rare. More often then not, we are challenged with evolving existing environment, with high expectations (but a tight budget).
So what should a generic design goal be (and feel free to keep this as a blueprint):
- Ensure maximum reuse of existing infrastructure, (off-course where possible).
Example: Switching infrastructure does not have the same change dynamics as Wi-Fi. It's not often that the buyer will be okay with replacing their couple of years old .af POE switches just to comply with the maximum potential of the new 802.11ac Wave 2 APs. The best you can hope for is adding a couple of more switches.
- Provide consistent and stable connection to as many clients as possible while not saturating the overall capacity of the ether.
Example: Overall use experience (speed, predictability, protocol overhead, device battery consumption, smaller CCI domain) is when the station to access point is close in proximity with strong RSSI and SNR.
- Have a defensive strategy for new technologies going in and using the same unlicensed spectrum (like the new LTE flavors operating, or to be operating, in unlicensed spectrum).
Example: What do you think will happen if you get an LTE eNodeB is positioned in the middle of the hallway, couple of feet away from your APs serving your high SLA clients?
- Defensive strategy to adapt to consistent changes in the client devices.
Example: This can really be anything. An office space that has been remodeled and people are sitting elsewhere, more devices per user, new "smart" devices, different and more demanding applications, implementing that new thing that everybody is talking about... IPv6 ;) and more.
- Policy: Ensure a consistent umbrella policy for your setup, yet ensure any changes are implemented quickly (with hopefully no downtime) and are implemented with a great degree of control.
So, who wants me to address the above 5 in more detail? What subject are you interested in mostly? Ask me on Twitter.