Wi-Fi Planning And The Need For Speed
This series of posts is dedicated to helping network managers plan a WLAN deployment by answering questions commonly asked when shopping for a long-term Wi-Fi solution.
With proper planning, you should have an accurate idea of the number and type of access points required for your environment, and if you are future-proofing your network, and anticipating a high number of mobile devices, then 802.11ac Wave2 is probably the way to go.
The 802.11ac Wi-Fi standard is a remarkable achievement for the WLAN industry, already moving beyond the Gigabit barrier. But for the mobile-first campus, is rolling out 802.11ac enough?
It’s certainly part of the answer, but remember you can go to your local PC store and pick up an 802.11ac access point for next to no cost, so what’s the difference between the SoHo and ‘enterprise’ WLAN solutions, if the advertised bandwidth is the same? The difference is how that bandwidth is managed and provisioned.
So what should I look for? Bandwidth management is achieved in various forms, but overall, you should be asking your vendor for the following:
1) Load Balancing & Band Steering
As an absolute minimum, WLAN solutions must be able to identify overcrowded access points or radios, and take action. Load balancing ensures that if an access point is carrying a high number of clients, and there is an underutilized access point nearby, then the clients can be redistributed and balanced across the two access points.
Band steering is a similar principle to load balancing, but it occurs within a single access point that has two radios. Commonly, mobile devices are equipped with both a 2.4GHz and 5GHz radio, and are programmed to favor one radio over the other, usually 5GHz for modern devices. That could leave a situation where many devices are connecting to the 5GHz radio, and only a handful on the 2.4GHz radio.
To learn more about band steering, read How band steering fixed a university Wi-Fi problem and Designing WLANS: What If We Could Double Our Airtime At 5 GHz?
Although more bandwidth is available on the 5GHz radio (there are more channels available to bond for 802.11n, and 802.11ac only operates on 5GHz), there are some situations where clients would be better served on the 2.4GHz radio, and that is when band steering will take charge and ensure that your two radio access point is being fully utilized.
2) Software Defined Radios
Following on from the above, the 2.4GHz spectrum is limited in capacity, and newer devices are pricing towards 5GHz. Because of spectrum congestion, you will undoubtedly face interference in a high capacity deployment.
To increase performance, it is actually recommended that you TURN OFF the 2.4GHz radio in 2/3 of your deployed access points. Recognizing the move away from 2.4GHz, but also knowing that 2.4GHz must still be supported for a time, some vendors have started to implement dual 5GHz radios within their access points, with the ability to configure one of the radios between 2.4GHz and 5GHz.
With this advancement, you can instantly deploy dual 5GHz access points in high capacity areas, then in the areas where you would otherwise disable the 2.4GHz radio, convert it to a 5GHz radio. This maximizes your investment both today and in the future, without needing to rip and replace devices.
3) L2/L3 Fast Secure Roaming
With more users and devices on the move, and many organizations enabling voice- and video-services over Wi-Fi, it is imperative that handoff between one access point to the next is seamless.
Most WLAN solutions on the market support fast roaming handoff within layer 2 domains. However if your access points are spread across multiple VLANs, then check how your vendor copes with (if at all) supporting a seamless handover between one VLAN to the next.
Typically, this will require a GRE tunnel between controllers or access points in the different VLANs that allow a client to maintain their original IP address until they have finished transmitting data. Then the tunnel will be torn down and the client will establish an IP address in the new VLAN.
Without Layer 3 roaming capabilities, the client may lose its connection, which could be highly problematic in areas where clients are on the border between the two VLANs.
4) Dynamic Airtime Scheduling & SLA’s
In a mixed client environment, older clients may slow down the performance of newer devices. SLA’s allow you to set a minimum targeted throughput level for certain device types, and if they are not met, a boost can be given through Dynamic Airtime Scheduling to push it up the priority queue.
Dynamic Airtime Scheduling also looks at the overall client landscape and can intelligently re-order the data transmissions of clients to improve the performance of newer devices such as .11ac/n, without actually impacting the performance of legacy .11g/a devices.
5) Context-Based QoS
While bandwidth optimization features such as band steering and load balancing maintain general order within your network, there are often user, device, or application groups that you want to prioritize, restrict, or even ban altogether.
Context-based access - which we will talk about in the security section later on in this series - enables the identification of your users, devices, and applications. Once identified, you can set different levels of access and service quality, for each.
For performance, this means that you can assign more bandwidth to your staff and students over guests; staff-owned devices over BYOD; or voice and video apps over gaming apps; it also allows the throttling or banning of illegal or bandwidth intensive apps including software updates and torrents.
You can learn more about this topic by reading How To Prioritize Devices On Your Network.
That's all for now. Next time we will discuss WLAN architecture, delving into the topic of cloud, controllers, and distributed control.All posts in this series: